JP6469041B2 - Control signal estimation device, pseudo force generation device, waveform estimation device, control signal estimation method, waveform estimation method, and program - Google Patents

Description

  The present invention relates to a technique for making a user perceive a pseudo force sense.

  There has been proposed a pseudo force sense generating device that controls a linear actuator based on a time-asymmetric rectangular wave control signal and perceives a pseudo force sense such as a traction illusion (see, for example, Non-Patent Document 1).

Tomohiro Amemiya, Shinya Takahata, Sho Ito, Hiroaki Gomi, “Brunavi 3”, a device that creates a sensation of being pulled by fingers, 2014, NTT Technical Journal, Vol. 26, No. 9, pp. 23-26 .

  Even when the same rectangular wave control signal is used, the perceived pseudo force sense varies greatly depending on the configuration and holding method of the pseudo force sense generator. However, in the conventional pseudo force sense generator, the control signal design method necessary to perceive the pseudo force sense is limited, and control signals suitable for the configuration and way of holding of each pseudo force sense device are provided. It was difficult to set.

  An object of the present invention is to estimate a target control signal for obtaining a target response waveform pattern that perceives a desired pseudo force sensation.

  A dynamic characteristic model of a pseudo force sensation generator to be modeled, which has a moving member that performs a periodic asymmetric motion according to a control signal waveform, and a gripper to which a force generated by the asymmetric motion is applied, and a gripper The relationship between the control signal waveform and the waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the skin obtained based on a plurality of models including the skin dynamic characteristic model in contact with the skin An estimated waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the skin obtained by applying the trial signal waveform as the control signal waveform to the forward dynamics model of the control target to be represented, and the target A trial signal waveform based on an error from a target response waveform pattern corresponding to at least one of the position, velocity, acceleration, and force applied to the skin Adjusted to obtain the target control signal waveform for controlling asymmetric movement.

  As described above, in the present invention, it is possible to estimate a target control signal for obtaining a target response waveform pattern that perceives a desired pseudo force sense.

FIG. 1 is a block diagram for explaining the configuration of the embodiment. 2A and 2B are conceptual diagrams for illustrating the pseudo force sense generating device of the embodiment. FIG. 3A is a conceptual diagram illustrating a state in which the pseudo force sense generating device is gripped with a finger, and FIG. 3B is a diagram of the dynamic characteristic model of the pseudo force sense generating device, the mechanical characteristic model of the skin, and the control signal waveform and the electric circuit. It is a conceptual diagram for demonstrating the electric circuit characteristic model showing the relationship with an output. FIG. 4A is a diagram illustrating a conventional rectangular wave control signal (control voltage). FIG. 4B is a diagram exemplifying a waveform of a force applied to the skin in contact with the grasping portion by the simulated force sense generating device driven based on the control signal of FIG. 4A. FIG. 4C is a diagram illustrating a waveform of the position of the grip portion of the pseudo force sense generator driven based on the control signal of FIG. 4A. FIG. 5A is a diagram illustrating a control signal (control voltage) according to the embodiment. FIG. 5B is a diagram exemplifying a waveform of force applied to the skin that is in contact with the gripping portion of the pseudo force sense generating device driven based on the control signal of FIG. 5A. FIG. 5C is a diagram illustrating a waveform of the position of the grip portion of the simulated force sense generator that is driven based on the control signal of FIG. 5A. FIG. 6A is a diagram illustrating a conventional rectangular wave control signal (control voltage). FIG. 6B is a diagram exemplifying a waveform of a force applied to the skin in contact with the grasping portion of the simulated force sense generating device driven based on the control signal of FIG. 6A. FIG. 6C is a diagram illustrating a waveform of the position of the grip portion of the pseudo force sense device driven based on the control signal of FIG. 6A. FIG. 7A is a diagram illustrating a control signal (control voltage) according to the embodiment. FIG. 7B is a diagram exemplifying a waveform of force applied to the skin in contact with the grasping portion of the pseudo force sense generator that is driven based on the control signal of FIG. 7A. FIG. 7C is a diagram illustrating a waveform of the position of the grip portion of the simulated force sense generator that is driven based on the control signal of FIG. 7A. FIG. 8A and FIG. 8B are conceptual diagrams for illustrating the simulated force sense generating device of the embodiment. FIG. 9 is a conceptual diagram for illustrating the pseudo force sense generation device of the embodiment. FIG. 10 is a conceptual diagram for explaining the dynamic characteristic model of the pseudo force sense generating device and the dynamic characteristic model of the skin. FIG. 11 is a block diagram illustrating a waveform estimation apparatus according to an embodiment that estimates a response waveform with respect to a control signal waveform. FIG. 12 is a block diagram for explaining the configuration of the embodiment. FIG. 13A is a diagram exemplifying the control signal (control voltage) of the embodiment. FIG. 13B is a diagram showing the force applied to the skin that the pseudo force sense generator driven based on the control signal of FIG. It is the figure which illustrated the waveform. FIG. 13C is a diagram exemplifying a waveform of the position of the grip portion of the simulated force sense generator driven based on the control signal of FIG. 13A. FIG. 14A is a diagram illustrating a control signal (control voltage) according to the embodiment. FIG. 14B is a diagram exemplifying a waveform of a force applied to the skin that is in contact with the gripping portion of the simulated force sense generator that is driven based on the control signal of FIG. 14A. FIG. 14C is a diagram exemplifying the waveform of the position of the grip portion of the simulated force sense generator driven based on the control signal of FIG. 14A. FIG. 15A to FIG. 15D are conceptual diagrams illustrating the configuration of the pseudo force sense generator according to the embodiment. 15B is a schematic plan view of the pseudo force generation device of the embodiment, FIG. 15A is a schematic cross-sectional view of 15A-15A in FIG. 15B, FIG. 15C is a schematic right side view of FIG. 15B, and FIG. It is a conceptual diagram for demonstrating the use condition of the pseudo force sense generator of form. 16A and 16B are conceptual diagrams for illustrating the vibrator according to the embodiment. FIG. 17A and FIG. 17B are conceptual diagrams illustrating the configuration of the simulated force sense generator according to the embodiment. FIG. 15A is a schematic plan view of the pseudo force generation apparatus of the embodiment, and FIG. 17B is a schematic right side view of FIG. 17A. FIG. 17C is a schematic cross-sectional view showing a modified example of the pseudo force sense generator. 18A to 18D are conceptual diagrams illustrating the configuration of the pseudo force sense generator according to the embodiment. 18A is a schematic plan view of the pseudo force sense generator according to the embodiment, FIG. 18B is a schematic right side view of FIG. 18A, and FIGS. 18C and 8D are schematic front views of FIG. 18A. FIG. 19A to FIG. 19C are conceptual diagrams illustrating the configuration of the pseudo force sense generator according to the embodiment. 19A is a schematic plan view of the pseudo force sense generator according to the embodiment, FIG. 19B is a schematic right side view of FIG. 19A, and FIG. 19C is a schematic bottom view of FIG. 19A. 20A to 20C are conceptual diagrams illustrating the configuration of the pseudo force sense generator according to the embodiment. 20A is a schematic plan view of the pseudo force sense generator according to the embodiment, FIG. 20B is a schematic right side view of FIG. 20A, and FIG. 20C is a schematic bottom view of FIG. 20A. FIG. 21A to FIG. 21C are conceptual diagrams illustrating the configuration of the pseudo force sense generator according to the embodiment. FIG. 21A is a schematic plan view of the pseudo force sense generator according to the embodiment, FIG. 21B is a schematic right side view of FIG. 21A, and FIG. 21C is a schematic front view of FIG. 22A and 22B are conceptual diagrams for illustrating the vibrator according to the embodiment. 23A and 23B are conceptual diagrams illustrating the configuration of the embodiment. FIG. 24 is a conceptual diagram illustrating the configuration of the embodiment. 25A to 25C are conceptual diagrams for explaining the operation of the modification. FIG. 26A to FIG. 26C are data illustrating characteristics of a conventional simulated force sense generator. FIG. 26A illustrates time-series data of the control voltage [V] of the simulated force sense generator, FIG. 26B illustrates time-series data of the force [N] applied to the skin from the simulated force sense generator, and FIG. The time series data of position [m] of a force sense generator are illustrated. The horizontal axis of FIGS. 26A to 26C represents time [sec]. 27A to 27C are data exemplifying characteristics of the simulated force sense generator according to the embodiment. 27A illustrates time-series data of the control voltage [V] of the simulated force sense generator, FIG. 27B illustrates time-series data of the force [N] applied to the skin from the simulated force sense generator, and FIG. The time series data of position [m] of a force sense generator are illustrated. The horizontal axis in FIGS. 27A to 27C represents time [sec]. FIG. 28A to FIG. 28C are data illustrating characteristics of the pseudo force sense generator according to the embodiment driven by the control voltage of the asymmetric rectangular waveform pattern. 28A illustrates time-series data of the control voltage [V] of the simulated force sense generator, FIG. 28B illustrates time-series data of the force [N] applied to the skin from the simulated force sense generator, and FIG. The time series data of position [m] of a force sense generator are illustrated. The horizontal axis of FIG. 28A-FIG. 28C represents time [sec]. 29A to 29C are data exemplifying characteristics of the pseudo force sense generator according to the embodiment in which an asymmetric rectangular waveform pattern is driven with a control voltage optimized as a target response waveform pattern. FIG. 29A illustrates time-series data of the control voltage [V] of the simulated force sense generator, FIG. 29B illustrates time-series data of force [N] applied to the skin from the simulated force sense generator, and FIG. The time series data of position [m] of a force sense generator are illustrated. The horizontal axis of FIGS. 29A to 29C represents time [sec]. FIG. 30A to FIG. 30C are data exemplifying characteristics of the simulated force sense generator according to the embodiment driven with a control voltage having an asymmetric rectangular waveform pattern. 30A illustrates time-series data of the control voltage [V] of the simulated force sense generator, FIG. 30B illustrates time-series data of the force [N] applied to the skin from the simulated force sense generator, and FIG. The time series data of position [m] of a force sense generator are illustrated. The horizontal axis of FIGS. 30A to 30C represents time [sec]. FIG. 31A to FIG. 31C are data exemplifying characteristics of the pseudo force sense generator according to the embodiment in which an asymmetric rectangular waveform pattern is driven with a control voltage optimized as a target response waveform pattern. FIG. 31A illustrates time-series data of the control voltage [V] of the simulated force sense generator, FIG. 31B illustrates time-series data of force [N] applied to the skin from the simulated force sense generator, and FIG. The time series data of position [m] of a force sense generator are illustrated. The horizontal axis of FIGS. 31A to 31C represents time [sec]. 32A and 32B show a period T1 in which the input waveform of the drive control signal is positive, a period T2 in which it is negative, and an asymmetry of the force applied to the skin from the pseudo force generation device (difference between positive and negative large forces) [N It is the figure which stem-plotted an example of the relationship with]. However, FIG. 32A illustrates the characteristics of the pseudo force sense generator of the embodiment driven by the control voltage of the asymmetric rectangular waveform pattern, and FIG. 32B is optimized using the asymmetric rectangular waveform pattern as the target response waveform pattern. The characteristic of the pseudo force sense generating device of the embodiment driven with the control voltage which is different is illustrated. FIG. 33A and FIG. 33B are diagrams illustrating time-series data of the control voltage [V] of the simulated force sense generator. FIG. 34 is a block diagram for explaining a modification of the pseudo force sense generating apparatus. FIG. 35 is a block diagram for explaining a modification of the pseudo force sense generating apparatus.

Embodiments of the present invention will be described below.
[Overview]
First, an outline will be described.
In the embodiment, a “pseudo force sense generator” having a “motion member” that performs periodic “asymmetric motion” according to a “control signal waveform” and a “gripping portion” to which a force generated by the asymmetric motion is applied Obtained from a plurality of models including a model of mechanical properties of the skin and a model of mechanical properties of the skin in contact with the “gripping unit”. “Waveform corresponding to at least one of stress” and “control signal waveform”, “reverse dynamics model to be controlled (for example, driving according to control signal when waveform corresponding to force applied to skin is given) "Model for obtaining the input side voltage of the" electric circuit "that supplies the signal)" to the "target response" corresponding to at least one of the position, speed, acceleration, and force applied to the skin Applying the waveform pattern ", obtain a" target control signal waveform "for controlling the" asymmetric movement ". The “inverse dynamics model to be controlled” is a model that considers at least both the “mechanical characteristic model of the pseudo force generation device” and the “mechanical characteristic model of the skin”. Therefore, a “target control signal waveform” necessary for realizing a desired “target response waveform pattern” can be easily obtained with various configurations and ways of holding the “pseudo force sense generator”. This makes it possible to realize a desired “target response waveform pattern” regardless of the configuration of the “pseudo force sense generator” and how it is held. In addition, since the “inverse dynamics model to be controlled” can be obtained in advance, the “target control signal waveform” can be obtained in real time with respect to the input “target response waveform pattern”. The “model” means a mathematical expression that represents a relationship between a plurality of values. In addition, “at least one of the position, speed, acceleration, and force applied to the skin of the grip portion” may be any one of these, or any one of them. The “waveform corresponding to at least one of the position, velocity, acceleration and the force applied to the skin of the grip portion” is a time waveform of “at least one of the position, velocity, acceleration and the force applied to the skin of the grip portion”. There may be a time waveform of the function value. “Control signal waveform” means a waveform corresponding to a control signal. For example, the “control signal waveform” may be the control signal itself, the input side voltage of the “electric circuit” that supplies a drive signal corresponding to the control signal, or the output side voltage (drive voltage). ), An input current, or an output current (drive current). “Target control signal waveform” means a waveform corresponding to a target control signal. Examples of the “target control signal waveform” include a voltage time waveform and a current time waveform. The time waveform of a certain physical quantity means a waveform made up of physical quantities at each time.

  The “inverse dynamics model to be controlled” may be a model obtained based on the “mechanical characteristic model of the pseudo force sense generator” and the “mechanical characteristic model of the skin”, or the “dynamics of the pseudo force sense generator”. Obtained based on the “characteristic model”, “skin dynamic characteristic model”, and “electric circuit characteristic model” representing the relationship between the control signal waveform and the output of the electric circuit that supplies the drive signal to the “pseudo force sense generator” It may be a model. By considering the “electric circuit characteristic model” like the latter, an appropriate “target control signal waveform” can be obtained according to the configuration of the electric circuit that supplies the drive signal to the “pseudo force sense generator”, It is possible to easily obtain a “target control signal waveform” necessary for realizing a desired “target response waveform pattern” with various “pseudo force sense generator” configurations and holding methods and “electric circuits”. An example of the “inverse dynamics model to be controlled” is the inverse of the mathematical model that obtains “a waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the skin” with respect to the “control signal waveform”. Function (inverse mapping) or approximate inverse function. The “waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the skin” represented by the “inverse dynamics model to be controlled” is the waveform corresponding to the “target response waveform pattern”. It is the same kind. For example, when the “reverse dynamics model to be controlled” represents the relationship between the “position of the gripper” and the “control signal waveform”, the waveform corresponding to the “target response waveform pattern” represents the waveform indicating the “position of the gripper” (Waveform representing position at each time). For example, when the “reverse dynamics model to be controlled” represents the relationship between the “gripping part speed” and the “control signal waveform”, the waveform corresponding to the “target response waveform pattern” represents the waveform representing the “gripping part speed”. (Waveform representing the speed at each time). For example, when the “reverse dynamics model to be controlled” represents the relationship between “force applied to the skin” and “control signal waveform”, the waveform corresponding to the “target response waveform pattern” represents the waveform representing “force applied to the skin”. (A waveform representing the force applied to the skin at each time).

  For example, the “pseudo force sense generating device” includes a “gripping part”, a “support part” whose relative position to the “gripping part” is fixed, an “elastic body” supported by the “support part”, and an “elastic body” ”And a“ motion member ”that performs a periodic asymmetric motion with respect to the“ support portion ”, and a force corresponding to the control signal waveform (for example, the output current or the output side voltage of the aforementioned“ electric circuit ”). And a coil provided to the “motion member”. In this case, an example of the “dynamic model of the pseudo force generation device” is the force applied to the “movement member” and the “gripping part” by the “coil”, and the position, velocity, and acceleration corresponding to the “gripping part”. And a mathematical expression that models the relationship between the position, velocity, and acceleration corresponding to the “movement member”. An example of the “skin dynamic system model” is a mathematical expression that models the relationship between the position and speed corresponding to the “grip” and the force applied to the skin.

  The “target response waveform pattern” represents, for example, a force whose direction is a predetermined direction (first direction) and whose absolute value is greater than or equal to a first threshold in the “first time interval”. In a “second time interval” different from the “interval”, a force whose direction is the direction opposite to the predetermined direction (second direction) and whose absolute value is within the first threshold (for example, less than the first threshold). Represent. The “first time interval” is shorter than the “second time interval”. That is, the force waveform in the “first time interval” and the force or acceleration waveform in the “second time interval” are asymmetric. Such a “target response waveform pattern” can present a clear pseudo force sense. Here, the “first threshold value” is preferably an absolute value of a lower limit value of a force that can be perceived by a person (note that there is an individual difference in the magnitude of a force that can be perceived, so a general lower limit value). It is also possible to obtain a value in advance and use that value). The force in the “second time interval” is preferably in the vicinity of the “first threshold value” within the “first threshold value”. The maximum absolute value of the force in the “first time interval” may be larger than the maximum absolute value of the force in the “second time interval”. For example, in the “first time interval” The maximum value of the absolute value of the force may be larger than the maximum value of the absolute value of the force in the “second time interval”. In particular, it is desirable that the “target response waveform pattern” is periodic, and that one period consists of a “first time interval” and a “second time interval” following the “first time interval”. In other words, it is desirable that one period of the “target response waveform pattern” is composed of only two time intervals of “first time interval” and “second time interval”. The “first time interval” included in one cycle is a continuous time interval. Similarly, the “second time interval” included in one cycle is a continuous time interval. Thereby, a clearer pseudo force sense can be presented. More preferably, the “target response waveform pattern” is an asymmetric rectangular waveform pattern (time asymmetric square wave) or an asymmetrical substantially rectangular waveform pattern. Thereby, a clearer haptic sense can be presented. In addition, “pseudo force sensation” means that a force that seems to move in the translational direction works even though the object (pseudo force sensation generator) does not actually perform a continuous translational movement in one direction. Perception that feels like this. The principle of the pseudo force sense is disclosed in Non-Patent Document 1, Reference Document 1, “Patent No. 4551448”, and the like. That is, consider the translational motion of an object with a certain mass. This translational motion is a periodic motion with partial acceleration (asymmetrical acceleration) that moves in a short time with a large acceleration in the direction in which the pseudo force sense is to be presented, and moves in a long time with a small acceleration in the opposite direction. Shall. In this case, the user holding the system including the object perceives a pseudo force sense in the presenting direction. This utilizes human perceptual characteristics, and is a phenomenon that occurs due to a peculiar sensation and a tactile sensation related to a gripping action. When presenting a pseudo force sense, the motion pattern of an object with a certain mass is not limited to “asymmetric acceleration”, but may be a motion in which the force applied to the skin is an asymmetric force pattern. It ’s fine. Furthermore, a first threshold value for the first time interval (a lower limit value of a size that can be perceived by a person or a value indicating a larger force) and a second threshold value for the second time interval (a person's It is also possible to individually set a lower limit value of the perceivable force magnitude or a value indicating a smaller force).

  The above-mentioned “target control signal waveform” is used as a control signal waveform to drive the “pseudo force sense generator”, and the “motion member” is a periodic “asymmetric motion” corresponding to the control signal waveform (target control signal waveform). To achieve the “target response waveform pattern” or its approximate response waveform pattern corresponding to at least one of the position, speed, acceleration of the “grip” and the force applied to the skin that touches the “grip” it can. As a result, it is possible to make the skin in contact with the “gripping part” perceive a clear simulated force sense. For example, the “pseudo force sense generating device” has a “gripping part”, a “support part” whose relative position to the “gripping part” is fixed, an “elastic body” supported by the “support part”, and an “elastic body” ”And a“ motion member ”that periodically performs“ asymmetric motion ”with respect to the“ support portion ”, and a coil that applies a force corresponding to the control signal waveform to the“ motion member ”. "Force" based on "is applied to the skin in contact with the grasping portion. This “force” waveform is periodic, and one cycle of the “force” waveform is composed of a “first time interval” and a “second time interval” following the “first time interval”. The direction of “force” in “first time interval” is “first direction”, and the direction of “force” in “second time interval” is “direction opposite to“ first direction ”. “Second direction”, and the force waveform in the “first time interval” and the force waveform in the “second time interval” are asymmetric. As a result, the skin in contact with the “grip” can be made to perceive a clear simulated force sense.

  "Grip position, velocity, acceleration, and force applied to the skin, obtained based on a plurality of models including the above-described" dynamic characteristic model of the pseudo force sense generator "and" skin dynamic characteristic model " The “target control signal waveform” may be applied as the “control signal waveform” to the “forward dynamics model to be controlled” representing the relationship between the “waveform corresponding to at least one of” and the “control signal waveform”. As a result, a response waveform corresponding to at least one of the position, speed, acceleration, and force applied to the skin when the asymmetric motion of the “movement member” is controlled by the “target control signal waveform” is estimated. it can. An example of “a forward dynamics model of a control target” is a mathematical model that obtains “a waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the skin” with respect to the “control signal waveform”. . The “forward dynamics model of the control target” may be a model obtained based on the “dynamic characteristic model of the pseudo force sense generator” and the “mechanical characteristic model of the skin”, or the “dynamics of the pseudo force sense generator”. It may be a model obtained on the basis of “characteristic model”, “skin dynamic characteristic model”, and “electric circuit characteristic model” representing the relationship between the control signal waveform and the output of the above “electric circuit”.

  It corresponds to at least one of the position, velocity, acceleration, and force applied to the skin obtained by applying the “trial signal waveform” as the “control signal waveform” to the “forward dynamic model of the control target” described above. Based on the “error” between the “estimated waveform” and the “target response waveform pattern” corresponding to at least one of the position, velocity, acceleration, and the force applied to the skin, the “trial signal waveform” The “target control signal waveform” for adjusting and controlling the asymmetric motion may be obtained. In other words, the “trial signal waveform” after such adjustment may be set as the “target control signal waveform”. The “trial signal waveform” may be adjusted so that the “error” is minimized, or the “trial signal waveform” may be adjusted so that the “error” is small. Even with this method, it is possible to easily obtain a “target control signal waveform” necessary for realizing a desired “target response waveform pattern” with various configurations and ways of holding “pseudo force sense generators”. As a result, a desired “target response waveform pattern” can be realized regardless of the configuration of the “pseudo force sensation generator” and how it is held. In addition, when the relationship between the “waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the skin and the waveform of the control signal” and the “control signal waveform” is non-linear, “from the forward dynamics model of the control target” to “ It may be difficult to obtain an “inverse dynamics model of the controlled object”. On the other hand, in the method of obtaining the “target control signal waveform” by adjusting so that the error between the “estimated waveform” and the “target response waveform pattern” is minimized, even if the above relationship is nonlinear, the nonlinear characteristic Thus, a “target control signal waveform” that takes this nonlinear characteristic into consideration can be obtained.

  Used by the “pseudo force generation device” modeled by the “reverse dynamics model of the control target” or “forward dynamics model of the control target” and the “reverse dynamics model of the control target” or “the forward dynamics model of the control target” The “pseudo force sensation generator” that generates the “target control signal waveform” may be the same (that is, using the parameters of the mechanical characteristics of the pseudo force sensation generator that is the actual machine to be controlled). “An inverse dynamics model of the controlled object” or “a forward dynamics model of the controlled object” may be obtained, or may not be the same (the parameters of the dynamic characteristics of the pseudo force sense generator other than the actual machine are used). Model may be obtained). In the former case, a more accurate “target control signal waveform” can be obtained with respect to a desired “target response waveform pattern”. In the latter case, it can be used to obtain the “target control signal waveform” of the “pseudo force sense generator” having the same or similar configuration as the modeled “pseudo force sense generator”. "Waveform" is highly versatile. That is, the obtained “target control signal waveform” is converted into a “pseudo force sensation generator” other than the “pseudo force sense generator” obtained from the parameters of the “reverse dynamics model to be controlled” or the “forward dynamics model to be controlled”. Can be applied.

  Using the “target control signal waveform” obtained as described above, the “base mechanism” supported directly or indirectly by the “first skin”, and the “target control signal waveform” with respect to the “base mechanism” A “pseudo force sensation generator” having a “contact mechanism” that applies a force based on the asymmetric motion to the “second skin” that directly or indirectly contacts the second skin It may be driven. However, the mass of the “contact mechanism” is smaller than the mass of the “base mechanism”, or the mass of the “contact mechanism” is less than the sum of the mass of the “base mechanism” and the mass of the mechanism attached to the “base mechanism”. small. In such a configuration, even when the mass of the “pseudo force generation device” as a whole is large, the mass of the “contact mechanism” is small relative to the “base mechanism” or “base mechanism” and the mechanism attached thereto. A sufficiently large force is transmitted from the “contact mechanism” to the skin. Thereby, even if the stroke of the asymmetrical motion of the “contact mechanism” is the same as that of the linear actuator of the “pseudo force sense generator” of the conventional method, a force sense can be presented more clearly. Alternatively, even if the stroke of the asymmetric motion of the “contact mechanism” is smaller than that of the conventional one, it is possible to present a force sense similar to that of the conventional method. That is, with such a configuration, a force sense can be presented more efficiently than in the past. Preferably, the mass of the “contact mechanism” is larger than zero and is not more than one third of the mass of the “base mechanism”. Thereby, a force sense can be presented more efficiently.

  The “target control signal waveform” having such a configuration is the “second base mechanism” and the “second base mechanism” that performs “second asymmetric motion” periodically according to the control signal waveform. Obtained based on a plurality of models including a mechanical characteristic model of a “second pseudo force generation device” having a “contact mechanism” and a mechanical characteristic model of skin contacting the “second contact mechanism”. The target “second contact mechanism” is added to the inverse dynamics model of the control target representing the relationship between the waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the skin and the control signal waveform. ”Is a waveform obtained by applying a“ target response waveform pattern ”corresponding to at least one of position, velocity, acceleration, and force applied to the skin. Preferably, the “inverse dynamics model to be controlled” includes the dynamic characteristic model of the “second pseudo force generation device”, the dynamic characteristic model of the skin, and the control signal waveform and the drive signal to the “second pseudo force generation device”. It is desirable to be obtained on the basis of an “electric circuit characteristic model” that represents the relationship with the output of the electric circuit that supplies the voltage. Alternatively, the “target control signal waveform” includes “second base mechanism” and “second contact mechanism” that performs “second asymmetric motion” periodically according to the control signal waveform with respect to “second base mechanism”. The second characteristic mechanism is obtained on the basis of a plurality of models including a dynamic characteristic model of the “second pseudo force generation device” having the above and a dynamic characteristic model of the skin in contact with the “second contact mechanism”. Obtained by applying the trial signal waveform as the control signal waveform to the forward dynamics model of the controlled object that represents the relationship between the waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the skin and the control signal waveform. And the estimated waveform corresponding to at least one of the position, velocity, acceleration and force applied to the skin, and the position, velocity, acceleration, and skin of the target “second contact mechanism”. Give Force may be a waveform obtained by adjusting a trial signal waveform based on an error between the "target response waveform pattern" corresponding to at least one. Preferably, the “forward dynamics model to be controlled” includes the dynamic characteristic model of the “second pseudo force generation device”, the dynamic characteristic model of the skin, and the control signal waveform and the drive signal to the “second pseudo force generation device”. It is desirable to be obtained on the basis of an “electric circuit characteristic model” that represents the relationship with the output of the electric circuit that supplies the voltage. The “second pseudo force generation device” is, for example, the “pseudo force sense generation device” itself driven by the “target control signal waveform”, and the “second base mechanism” is the “base mechanism” and the “second contact mechanism”. "Contact mechanism", "second asymmetric motion" may be "asymmetric motion", or other "pseudo force sense generator". Even if “second base mechanism” is not “base mechanism”, the difference between the mass of “second base mechanism” and the mass of “base mechanism” is 10% of the mass of “base mechanism” or “second base mechanism”. It may be less than or equal to the degree. Even if the “second contact mechanism” is not the “contact mechanism”, the difference between the mass of the “second contact mechanism” and the mass of the “contact mechanism” is equal to the mass of the “contact mechanism” or the “second contact mechanism”. What is necessary is just about 10% or less.

  The “periodic asymmetric motion” is a vibration for causing a pseudo force sense to be perceived by the force applied to the skin from the “contact mechanism”, and the time-series waveform of the vibration in the “predetermined direction” This means a vibration that is asymmetric with the time-series waveform of vibration in the opposite direction. An example of “periodic asymmetric motion” is asymmetric vibration. An example of “periodic asymmetric motion” is “asymmetric vibration”. “Asymmetric vibration” is, for example, a time-series waveform of “physical quantity” of “predetermined direction” of “contact mechanism” with respect to “base mechanism” and a direction opposite to “predetermined direction” of “contact mechanism” with respect to “base mechanism”. The vibration of the “contact mechanism” is asymmetric with the time-series waveform of “physical quantity”. Examples of “physical quantities” include acceleration, speed, or position of the “contact mechanism” relative to the “base mechanism”, the force that the “contact mechanism” applies to the “base mechanism”, the force that is applied to the skin from the “contact mechanism”, or “ Acceleration, speed, or position of the “contact portion”.

  The “base mechanism” is directly or indirectly supported by the “first skin”, and the “contact mechanism” applies a force based on asymmetric motion to the “second skin” in direct or indirect contact. In this case, the “asymmetric motion” may be “asymmetric vibration” corresponding to the “target control signal waveform” and the “bias component”. The “bias component” is a component that controls the vibration center of “asymmetric vibration”, and is a component that moves the vibration center in the direction in which a pseudo force sense is presented due to “asymmetric vibration”, for example.

  The “first skin” and the “second skin” are preferably the same target (animal including a human), but the “first skin” and the “second skin” are different. It may be. “Α directly or indirectly supported by β” means α directly supported by β or α supported by β via an inclusion. “Α is supported by β” means that part or all of the motion of α is restricted by β, for example, part or all of the freedom of motion of α is restricted by β. It is that. Not only when α is fixed to β, but even if α can move or rotate relative to β, if α is limited by β, if “α is β It can be said that “Directly or indirectly contacted β” may be β that is contacted without inclusions, or may be β that is contacted via inclusions. The inclusion may be a rigid body, an elastic body, a plastic body, a fluid, or a combination of at least some of the characteristics, but can transmit the force from the “contact portion” to the skin. There is a need.

  In addition, the “contact mechanism” may be driven based on a control signal waveform different from the “target control signal waveform”. That is, a “base mechanism” that is directly or indirectly supported by the “first skin” and a periodic asymmetric motion for causing the “base mechanism” to perceive a pseudo force sense, Alternatively, it may be a “pseudo force sense generating device” having a “contact mechanism” that applies a force based on asymmetric motion to the “second skin” that is indirectly contacted. However, the mass of the “contact mechanism” is smaller than the mass of the “base mechanism”.

[First Embodiment]
The first embodiment will be described below with reference to the drawings.
<Configuration>
As illustrated in FIG. 1, the system of this embodiment includes a control signal estimation device 11 that obtains a target control signal waveform from a target response waveform pattern, a pseudo force sense generation device 12 that presents a pseudo force sense, and a control signal estimation device 11. An electric circuit 13 is provided for obtaining a drive signal of the pseudo force sense generator 12 (actuator) from the obtained target control signal waveform. The control signal estimation device 11 includes a storage unit 111, an input unit 112, and a model application unit 113.

<Control signal estimation device 11>
The control signal estimation device 11 includes, for example, a general-purpose or dedicated memory including a processor (hardware processor) such as a CPU (central processing unit) and a memory such as random-access memory (RAM) and read-only memory (ROM). An apparatus configured by a computer executing a predetermined program. The computer may include a single processor and memory, or may include a plurality of processors and memory. This program may be installed in a computer, or may be recorded in a ROM or the like in advance. In addition, some or all of the processing units are configured using an electronic circuit that realizes a processing function without using a program, instead of an electronic circuit (circuitry) that realizes a functional configuration by reading a program like a CPU. May be. In addition, an electronic circuit constituting one device may include a plurality of CPUs.

<Electric circuit 13>
The electric circuit 13 is a circuit having characteristics such as a high-pass filter.

<Pseudo force sense generator 12>
The simulated force sense generator 12 includes, for example, a support part 121, springs 122 and 123 (elastic body), a coil 124, a motion member 125 that is a permanent magnet, and a grip part 126 (case). The gripping part 126 and the support part 121 of this embodiment are both hollow members having a shape in which both open ends of a cylinder (for example, a cylinder or a polygonal cylinder) are closed. However, the support portion 121 is smaller than the grip portion 126 and has a size that can be accommodated in the grip portion 126. The gripping part 126 and the support part 121 are made of synthetic resin such as ABS resin, for example. The springs 122 and 123 are, for example, a helical spring or a leaf spring made of metal or the like. The elastic coefficients (spring constants) of the springs 122 and 123 are preferably the same, but may be different from each other. The motion member 125 is, for example, a cylindrical permanent magnet, and one end 125a side in the longitudinal direction is an N pole, and the other end 125b side is an S pole. The coil 124 is, for example, a continuous enamel wire, and includes a first winding portion 124a and a second winding portion 124b.

  The movement member 125 is accommodated in the support portion 121 and is supported so as to be slidable in the longitudinal direction. Although details of such a support mechanism are not shown in the drawings, for example, a straight rail along the longitudinal direction is provided on the inner wall surface of the support portion 121, and a rail support portion that slidably supports the rail on the side surface of the motion member 125. Is provided. One end of the spring 122 is fixed to the inner wall surface 121 a on one end side in the longitudinal direction of the support portion 121 (that is, one end of the spring 122 is supported by the support portion 121), and the other end of the spring 122 is the end of the motion member 125. It is fixed to the portion 125a (that is, the end portion 125a of the moving member 125 is supported by the other end of the spring 122). Further, one end of a spring 123 is fixed to the inner wall surface 121b on the other end side in the longitudinal direction of the support portion 121 (that is, one end of the spring 123 is supported by the support portion 121), and the other end of the spring 123 is an exercise member. 125 is fixed to an end portion 125b of 125 (that is, the end portion 125b of the moving member 125 is supported by the other end of the spring 123).

A coil 124 is wound around the outer peripheral side of the support portion 121. However, the end portion 125a side of the moving member 125 (N pole side), and the first winding portion 124a is wound in the _{A 1} direction (direction toward the back to the front), the end portion 125b side (S-pole side in), a second winding portion 124b is wound in the a ₁ direction in the opposite direction to the direction of B ₁ direction (direction toward the front to the back). That is, when viewed from the end 125a side (N pole side) of the moving member 125, the first winding portion 124a is wound clockwise and the second winding portion 124b is wound counterclockwise. Further, in a state where the motion member 125 is stopped and the elastic forces from the springs 122 and 123 are balanced, the end portion 125a side (N pole side) of the motion member 125 is disposed in the region of the first winding portion 124a, and the end portion It is desirable that the 125b side (S pole side) be disposed in the region of the second winding portion 124b.

  The support portion 121, the springs 122 and 123, the coil 124, and the motion member 125 arranged and configured as described above are accommodated in the grip portion 126, and the support portion 121 is fixed inside the grip portion 126. That is, the relative position of the grip part 126 with respect to the support part 121 is fixed. However, the longitudinal direction of the grip portion 126 coincides with the longitudinal direction of the support portion 121 and the longitudinal direction of the motion member 125.

The coil 124 applies a force corresponding to the flowed current to the motion member 125, whereby the motion member 125 is periodically asymmetrically moved with respect to the support portion 121 (asymmetric in the axial direction with respect to the support portion 121. Periodic translational reciprocating motion). That, S when an electric current is applied to the _{A 1} direction _{(B 1} direction) to the coil 124, by reaction of the Lorentz force is described in Fleming's left-hand rule, the _{C 1} direction (the movement member 125 N pole to the motion member 125 A force in the direction toward the pole (right direction) is applied (FIG. 2A). Conversely, when an electric current is applied to the _{A 2} direction _{(B 2} direction) to the coil 124, the motion member 125 _{C 2} direction (direction toward the S pole of the moving member 125 to the N pole: leftward) force is applied ( FIG. 2B). However, _{A 2} direction is opposite the direction of _{A 1} direction. By these operations, kinetic energy is given to the system including the moving member 125 and the springs 122 and 123. Thereby, the position and acceleration (the position and acceleration in the axial direction with reference to the support portion 121) of the motion member 125 with respect to the grip portion 126 can be changed.

<Pretreatment>
In the preprocessing of the present embodiment, the inverse dynamics model Mc to be controlled is set and stored in the storage unit 111. The inverse dynamics model Mc to be controlled according to the present embodiment corresponds to a waveform corresponding to at least one of the position, speed, and acceleration of the gripping portion 126 of the pseudo force sense generator 12 and the force applied to the skin contacting the gripping portion 126. 4 is a mathematical model for obtaining a waveform based on a waveform of a control signal (control signal waveform) for driving the moving member 125. The inverse dynamics model Mc to be controlled according to the present embodiment is obtained based on the dynamic characteristic model Md of the pseudo force sense generator 12 and the dynamic characteristic model Ms of the skin in contact with the grasping portion 126.

  As illustrated in FIG. 3A, a state is assumed in which the outside of the grip portion 126 of the pseudo force sense generation device 12 of the present embodiment is gripped by the user's finger 100. 3A illustrates a state in which the side surface side of the grip portion 126 (outside the side surface of the pseudo force sense generator 12) is gripped, the end surface side of the grip portion 126 (the longitudinal length of the pseudo force sense generator device 12). A state in which the outer side of the direction end face) is gripped may be assumed.

As illustrated in FIG. 3B, this state is expressed by a dynamic characteristic model Md of the pseudo force sense generator 12 and a dynamic characteristic model Ms of the skin of the finger 100 that is in contact with (holding) the grasping unit 126. The dynamic characteristic model Md of the pseudo force generation device 12 of this example includes mass points M ₁ and M _{2 having} masses m ₁ and m ₂ , a spring having an elastic coefficient k ₂ connecting them, and a viscosity coefficient (damping coefficient). b ₂ of the damper, and represent the dynamics characteristics of comprising periodic Lorentz force f acting on the mass point _M 1, _{M 2} according to the drive voltage Vout. In the case of the configuration illustrated in FIGS. 2A and 2B, the Lorentz force f can be expressed as f = i ₂ BL. However, i ₂ [A] is a current flowing through the coil 124, B is a magnetic flux density by the coil 124, and L [m] is a length of the coil 124 perpendicular to the magnetic flux direction penetrating the support portion 121 in the longitudinal direction. It is. The position of the mass point _{M 1} with respect to the reference origin _{O 1} is expressed as _{x 1,} the position of the mass point _{M 2} with respect to the reference origin _{O 2} expressed as _{x 2.} However, the reference origins O ₁ and O ₂ are points at which relative positions with respect to the center of gravity of the finger 100 are fixed. Further, in both x ₁ and x ₂ , the right direction in FIG. 3A from the center of gravity of the finger 100 is positive, and the left direction in FIG. 3A from the center of gravity of the finger 100 is negative. Here, it is assumed that the center of gravity of the finger does not move with respect to the outside world. The time differential value of x ₁ and x ₂ , that is, the speed

Is written. However, in the present specification, these may be expressed as x ^· ₁ , x ^· ₂ due to restrictions on the description. A skin dynamic characteristic model Ms illustrated in FIG. 3B represents characteristics of a dynamic system including a spring having an elastic coefficient k ₁ and a damper having a viscosity coefficient b ₁ between the mass point M ₁ and the center of gravity of the finger 100. Here, the force (for example, stress) applied to the skin of the finger 100 in contact with the grip portion 126 is expressed as fs.

The mathematical expression of the dynamic characteristic model Md of the simulated force sense generator 12 and the dynamic characteristic model Ms of the skin is as follows, for example.
≪Example of mechanical property model Md≫

The dynamic system parameters m ₁ , m ₂ , k ₂ , and b ₂ of the mechanical characteristic model Md may be obtained from design values or measured values of the pseudo force sense generator 12 or obtained by a method such as system identification. Also good.
≪Example of mechanical property model Ms≫
fs = k ₁ · x ₁ + b ₁ · x ^· ₁ (2)
The dynamic system parameters k ₁ and b ₁ of the dynamic characteristic model Ms may be obtained by a method such as system identification or may be a typical value.

<< Example of inverse dynamics model Mc to be controlled >>
The unknown time series parameters in the above equations (1) and (2) are f, x ₁ , x ^· ₁ , x ₂ , x ^· ₂ and fs. By erasing ₁ , x ^· ₁ , x ₂ , x ^· ₂ , the relational expression fs = F (f) between f and fs is obtained. Here, in the example of FIGS. 2A and 2B, it can be expressed as f = i ₂ BL. B and L are obtained by a method such as a design value or system identification of the pseudo force sense generator 12. From fs = F (i ₂ BL) = F ₂ (i ₂ ), the relational expression fs = F ₂ (i ₂ ) (3A)
Is obtained. I ₂ = Inv (fs) (3B) which is an inverse function or an approximate inverse function of the relational expression (3A)
May be the inverse dynamics model Mc to be controlled.

Assuming that the resistance value of the coil 124 is R and the voltage applied to the coil 124 is Vout, i ₂ = Vout / R, so fs = F ₂ (Vout / R) = F _R (Vout). This relational expression fs = F _R (Vout) (4A)
Vout = Inv _R (fs) (4B) which is an inverse function or an approximate inverse function of
May be the inverse dynamics model Mc to be controlled.

From the above equation (1) (2), the relational expression to obtain the acceleration ^{x · ·} ₁ giving the skin of the finger 100 in contact with the grip portion 126 with respect to _{^{i 2 x ·· 1 = F m}} (i 2) (5A )
And i ₂ = Inv _m (x ^· ₁ ) (5B) which is an inverse function or an approximate inverse function of the relational expression (5A).
May be the inverse dynamics model Mc to be controlled.

From the above formulas (1) and (2), relational expression x ^·· ₁ = F _Rm (Vout) (6A) for obtaining acceleration x ^·· ₁ with respect to Vout
And Vout = Inv _Rm (x ^·· ₁ ) (6B) which is an inverse function or an approximate inverse function of the relational expression (6A).
May be the inverse dynamics model Mc to be controlled.

From Expression (1) and Expression (2), a relational expression x ^· ₁ = F ′ (f) for obtaining x ^· ₁ for f (7A)
And i ₂ = Inv (x ^· ₁ ) (7B) which is an inverse function or an approximate inverse function of the relational expression (7A).
May be the inverse dynamics model Mc to be controlled.

From Expression (1), Expression (2), and i ₂ = Vout / R, a relational expression x ^· ₁ = F′x _{· R} (Vout) for obtaining x ^· ₁ with respect to Vout (8A)
Vout = Inv _{x · R} (x ^· ₁ ) (8B) which is an inverse function or an approximate inverse function of the relational expression (8A)
May be the inverse dynamics model Mc to be controlled.

From Expression (1) and Expression (2), relational expression x ₁ = F _x (f) (9A) for obtaining x ₁ for f
And i ₂ = Inv _x (x ₁ ) (9B) which is an inverse function or an approximate inverse function of the relational expression (9A).
May be the inverse dynamics model Mc to be controlled.

From Expression (1) and Expression (2), relational expression x ₁ = F _V (Vout) (10A) for obtaining x ₁ with respect to Vout
And Vout = Inv _R (x ₁ ) (10B) which is an inverse function or an approximate inverse function of the relational expression (10A).
May be the inverse dynamics model Mc to be controlled.

<Target control signal waveform estimation process>
In the target control signal waveform estimation process of this embodiment, a target response waveform pattern is input to the input unit 112. The target response waveform pattern corresponds to at least one of the position, speed, acceleration, and force of the finger 100 applied to the skin of the target gripper 126. Which physical quantity time waveform pattern (a waveform pattern made up of physical quantities at each time) is set as the target response waveform pattern depends on the configuration of the inverse dynamics model Mc to be controlled. That is, the control target inverse dynamics model Mc is a model for obtaining a target control signal waveform from the target response waveform pattern, and the target response waveform pattern input to the input unit 112 is a time waveform corresponding to the control target inverse dynamics model Mc. It becomes a pattern. For example, in the case of the inverse dynamics model Mc to be controlled in Expression (3B) or Expression (4B), the time waveform pattern of the force fs is set as the target response waveform pattern. In the case of the inverse dynamics model Mc to be controlled of Expression (5B) or Expression (6B), the time waveform pattern of the acceleration x ^·· ₁ is set as the target response waveform pattern. In the case of the inverse dynamics model Mc to be controlled of Expression (7B) or Expression (8B), the time waveform pattern of speed x ^· ₁ is set as the target response waveform pattern. The target response waveform pattern time waveform patterns of the position x ₁ in the case of inverse dynamics model Mc of the controlled object of Formula (9B) or formula (10B). The target response waveform pattern is a pattern of a control signal for periodically asymmetrically moving the moving member 125 of the pseudo force sense generator 12. For example, the target response waveform pattern represents a force whose direction is a predetermined direction and whose absolute value is equal to or greater than a first threshold in the first time interval T1, and is different from the first time interval T1. T2 is a periodic waveform pattern representing a force whose direction is opposite to the predetermined direction and whose absolute value is within the second threshold. The first time interval T1 is shorter than the second time interval T2 (T1 <T2). For example, one cycle of this cycle consists of T1 and T2. Preferably, the target response waveform pattern is a rectangular waveform pattern or a substantially rectangular waveform pattern.

The target response waveform pattern is sent to the model application unit 113. The model application unit 113 applies the target response waveform pattern to the inverse dynamics model Mc to be controlled read from the storage unit 111, and obtains and outputs the target control signal waveform. That is, the model application unit 113 substitutes the physical quantity of each time of the target response waveform pattern into the inverse dynamics model Mc to be controlled, obtains a function value corresponding thereto, and outputs a time waveform of the function value. For example, in the case of equations (3B), (5B), and (9B), the time waveform of i ₂ is the target control signal waveform, and in the case of equations (4B), (6B), (8B), and (10B), Vout The time waveform is the target control signal waveform.

  A control signal (control voltage or control current) corresponding to the obtained target control signal waveform is applied to the electric circuit 13. The electric circuit 13 sends a corresponding drive signal to the pseudo force sense generator 12. The simulated force sense generator 12 is driven based on this, and presents a simulated force sense to the finger 100 gripping the grip portion 126.

<Features of this embodiment>
In this embodiment, the target control signal waveform for controlling the pseudo force sense generator 12 can be estimated from the target response waveform pattern for presenting the pseudo force sense. Thereby, even when the configuration and the way of holding the pseudo force sense generator 12 are changed, an appropriate target control signal waveform can be estimated according to each. Further, since the inverse dynamics model Mc to be controlled can be obtained in advance by an inverse function or an approximate inverse function, the target control signal waveform can be estimated in real time from the target response waveform pattern.

  In addition, the characteristics necessary for the amplifier and the characteristics necessary for the simulated force sense generator 12 that are necessary to approximately realize the target response waveform pattern can be known, and can be utilized in the design thereof. When the characteristics of the amplifier and the characteristics of the simulated force sense generator 12 are defined, it is possible to know a range in which an effective stimulus for generating a desired simulated force sense can be realized. When a modeling error is expected, it is possible to simulate how much output error is generated, which is useful for designing the dynamic system of the pseudo force sense generator 12.

  The mechanical characteristic models Md and Ms are obtained by modeling the state in which the grip portion 126 (that is, the grip portion of the actual machine) of the pseudo force sense device 12 controlled based on the target control signal waveform is gripped. Alternatively, it may be a model of a state in which the grip portion of another device that is the same as or similar to the simulated force sense generator 12 is gripped. In the former case, a more accurate target control signal waveform can be estimated.

[Second Embodiment]
The second embodiment is a modification of the first embodiment, and the inverse dynamics model Mc to be controlled is the dynamic characteristic model Md of the pseudo force sense generator 12, the dynamic characteristic model Ms of the skin, and the control signal waveform and the electric circuit. The difference is based on the electric circuit characteristic model Me representing the relationship with the output of. Below, it demonstrates centering around difference with 1st Embodiment, About the already demonstrated matter, description is simplified using the same reference number.

<Configuration>
As illustrated in FIG. 1, the system of the present embodiment includes a control signal estimation device 21 that obtains a target control signal waveform from a target response waveform pattern, a pseudo force sense generator 12, and an electric circuit 13. The control signal estimation device 21 includes a storage unit 211, an input unit 112, and a model application unit 213. The control signal estimation device 21 is a device configured by, for example, the above-described general-purpose or dedicated computer executing a predetermined program.

<Pretreatment>
In the preprocessing of the present embodiment, the inverse dynamics model Mc to be controlled is set and stored in the storage unit 211. The inverse dynamics model Mc to be controlled according to the present embodiment includes a waveform corresponding to at least one of the position, speed, acceleration of the gripping portion 126 of the simulated force sense generator 12 and the force applied to the skin that contacts the gripping portion 126, and It is a numerical model showing the relationship with the waveform (control signal waveform) of the control signal for driving the moving member 125. The inverse dynamics model Mc to be controlled in this embodiment is driven by the dynamic characteristic model Md of the simulated force sense generator 12, the dynamic characteristic model Ms of the skin that contacts the grasping portion 126, and the control signal waveform and the simulated force sense generator 12. It is obtained based on the electric circuit characteristic model Me representing the relationship with the output of the electric circuit that supplies the signal.

  As illustrated in FIG. 3A, a state is assumed in which the outside of the grip portion 126 of the pseudo force sense generation device 12 of the present embodiment is gripped by the user's finger 100. The control voltage Vin is input to the electric circuit of the transfer function H (S), the output of the transfer function H (S) becomes the drive voltage Vout, and the pseudo force sense generator 12 is driven based on the drive voltage Vout. S represents a Laplace transform variable.

  As illustrated in FIG. 3B, this state is represented by the dynamic characteristic model Md and the dynamic characteristic model Ms described in the first embodiment, the control signal waveform (for example, the input side voltage or input current of the electric circuit), and the output of the electric circuit. It is expressed by an electric circuit characteristic model Me representing the relationship between When modeling is performed in this way, it is possible to construct a forward dynamics model in which a force applied to m1 and m2 and fs are obtained when Vin is given. The output of the electric circuit is, for example, a waveform of a drive signal (drive voltage or drive current) for driving the moving member 125.

<< Example of electrical circuit characteristic model Me >>
The following relationship holds between the control voltage Vin and the drive voltage Vout.
Vout = H (S) · Vin (11)
However, H (S) = S · T / (1 + S · T), S is a variable of Laplace transform, and T is a time constant of the filter. These characteristics may be obtained from design values or may be obtained by a method such as system identification.

<< Example of inverse dynamics model Mc to be controlled >>
A relational expression based on the above-described expressions (1) and (2) and the above-described expression (11) is set as an inverse dynamics model Mc to be controlled. For example, a relational expression based on any one of the above formulas (3B) to (10B) and the formula (11) is set as the inverse dynamics model Mc to be controlled. For example, Vout = Inv _R (fs) = H (S) · Vin holds from Expression (4B) and Expression (11). Therefore, the following relational expression can be used as the inverse dynamics model Mc to be controlled.
Vin = H ⁻¹ (S) · Inv _R (fs) (12)
Or it is good also considering the following relational expressions obtained from Formula (6B) and Formula (11) as the inverse dynamics model Mc to be controlled.
Vin = H ⁻¹ (S) · Inv _R (x ^·· ₁ ) (13)
Alternatively, for example, it may be a reverse dynamics model Mc to be controlled using the following Vout ′ instead of Vout, indicating that Vout is saturated at ± 1.

<Target control signal waveform estimation process>
In the target control signal waveform estimation process of this embodiment, a target response waveform pattern is input to the input unit 112. For example, in the case of the inverse dynamics model Mc to be controlled of Expression (12), the time waveform pattern of the force fs is set as the target response waveform pattern. For example, in the case of the inverse dynamics model Mc to be controlled of Expression (13), the time waveform pattern of acceleration x ^·· ₁ is set as the target response waveform pattern.

  The target response waveform pattern is sent to the model application unit 113. The model application unit 113 applies the target response waveform pattern to the inverse dynamics model Mc to be controlled read from the storage unit 111, and obtains and outputs the target control signal waveform. In the case of the inverse dynamics model Mc to be controlled of Expressions (12) and (13), the time waveform of Vin is the target control signal waveform.

  A control signal (control voltage or control current) corresponding to the obtained target control signal waveform is applied to the electric circuit 13, and the electric circuit 13 sends a drive signal corresponding thereto to the simulated force sense generator 12. The simulated force sense generator 12 is driven based on this, and presents a simulated force sense to the finger 100 gripping the grip portion 126.

<Experimental result>
A comparison experiment result between the conventional method of Non-Patent Document 1 and this embodiment is illustrated.
<< Experimental result 1: [T1, T2] = [2,10] msec >>
4A to 4C show experimental results by the conventional method. As shown in FIG. 4A, in the conventional method, the control voltage [V] of the pseudo force sense generator 12 is a substantially rectangular wave of [T1, T2] = [2, 10] msec. That is, the time interval T1 of 2 [msec] and the time interval of 10 [msec] are alternately repeated, and the control voltage Vin in the time interval T1 is positive and the control voltage Vin in the time interval T2 is negative. The absolute value of the control voltage Vin in the time interval T1 is larger than the absolute value of the control voltage Vin in the time interval T2. FIG. 4B shows a waveform of the force [N] applied to the skin of the finger 100 in contact with the grip part 126 by the pseudo force generation device 12 controlled in this way, and FIG. 4C shows the position [m] of the grip part 12. Represents a waveform.

  5A to 5C show experimental results according to this embodiment. In this experiment, the inverse dynamics model Mc to be controlled was applied to the target response waveform pattern indicated by the broken line in FIG. 5B to obtain the control voltage waveform (control signal waveform) in FIG. 5A. This target response waveform pattern (FIG. 5B) is a time waveform of the force [N] applied to the skin of the finger 100 that contacts the grasping portion 126. This target response waveform pattern is a periodic pattern in which a time interval T1 of 2 [msec] and a subsequent time interval T2 of 10 [msec] are repeated, and the time interval T1 and the time interval T2 are 1 in one cycle. One period is 12 [msec] consisting of a time interval T1 and a time interval T2. This target response waveform pattern is a substantially rectangular waveform pattern, and represents a force whose direction is the first direction (positive) and whose absolute value is greater than or equal to the threshold Th1 in the time interval T1, and in the time interval T2, the first response The force is in the direction opposite to the direction (second direction: negative) and the absolute value is within the threshold value Th2 (Th2 <Th1). The waveform pattern in the time interval T1 and the waveform pattern in the time interval T2 are asymmetric (the force time waveform in the time interval T1 and the force time waveform in the time interval T2 are asymmetric). This is for presenting a pseudo force sense (traction force or the like) in the first direction. The response waveform of the force [N] applied to the skin of the finger 100 that is in contact with the grasping portion 126 by the simulated force sense generator 12 controlled by the control voltage (FIG. 5A) according to the target response waveform pattern is a solid line in FIG. It shows with. FIG. 5B shows that a response waveform that approximates the target response waveform pattern is obtained. FIG. 5C shows a waveform of the position [m] of the grip portion 12 at this time.

<< Experimental result 2: [T1, T2] = [3, 16] msec >>
6A to 6C show experimental results according to the conventional method, and FIGS. 7A to 7C show experimental results according to the present embodiment. Unlike Comparative Experiment Result 1, [T1, T2] = [3,16] msec. Other conditions are the same as those in Comparative Experiment Result 1.

  4B and FIG. 5B, FIG. 4C and FIG. 5C, FIG. 6B and FIG. 7B, and FIG. 6C and FIG. It can be seen that there is little fine vibration of the waveform, while the difference between the maximum value in the positive direction and the maximum value in the negative direction is large, and the asymmetry is large. Thereby, it turns out that a clear force sense can be shown in this form rather than the conventional method.

<Features of this embodiment>
Also in this embodiment, the target control signal waveform for controlling the pseudo force sense generator 12 can be estimated from the target response waveform pattern for presenting the pseudo force sense. Thereby, even when the configuration and the way of holding the pseudo force sense generator 12 are changed, an appropriate target control signal waveform can be estimated according to each. Further, since the inverse dynamics model Mc to be controlled can be obtained in advance by an inverse function or an approximate inverse function, the target control signal waveform can be estimated in real time from the target response waveform pattern. Further, since the inverse dynamics model Mc to be controlled is obtained in consideration of the electric circuit characteristic model Me, even when the characteristic of the electric circuit 13 changes, an appropriate target control signal waveform can be estimated according to each.

  The mechanical characteristic models Md and Ms are obtained by modeling the state in which the grip portion 126 (that is, the grip portion of the actual machine) of the pseudo force sense device 12 controlled based on the target control signal waveform is gripped. Alternatively, it may be a model of a state in which the grip portion of another device that is the same as or similar to the simulated force sense generator 12 is gripped. Similarly, the electric circuit characteristic model Me may be a model of the electric circuit 13 (that is, an actual machine) to which a control signal having a target control signal waveform is given, or other electric circuits that are the same as or similar to the electric circuit 13. It may be a model of a circuit. In either case, a more accurate target control signal waveform can be estimated in the former case.

[Modification 1 of the first and second embodiments]
As a modified example of the simulated force sense generator 12 (FIGS. 2A and 2B), the coil 124 is wound only on the end 125a side (N pole side) or the end portion 125b side (S pole side) of the motion member 125. Also good. For example, the coil 124 may be wound only on the end portion 125b side (S pole side) on the outer peripheral side of the support portion 121 as in the pseudo force sense generation device 12 ′ of FIGS. 8A and 8B. The other configuration of the pseudo force sense generator 12 ′ is the same as that of the pseudo force sense generator 12. The same reference numerals as those of the pseudo force sense generator 12 are used for parts common to the pseudo force sense generator 12 of the pseudo force sense generator 12 ′.

[Modification 2 of the first and second embodiments]
Or, instead of the coil 124 of the pseudo force sense generator 12 (FIGS. 2A and 2B) as in the pseudo force sense generator 12 ″ of FIG. 9, the end portion 125a side (N pole) on the outer peripheral side of the support portion 121 is used. a first winding portion 124a 'coil wound around the _{a 1} direction side), the end portion 125b side of the outer peripheral side of the supporting portion 121 (S pole side) in the _{B 1} direction _{(a 1} direction opposite) The coil on the second winding part 124b ′ side wound around the coil may be used independently, that is, the coil on the first winding part 124a ′ side and the coil on the second winding part 124b ′ side are electrically connected. The pseudo force sense generator 12 ″ may have the same configuration as that of the pseudo force sense generator 12. The same reference numerals as those of the pseudo force sense generator 12 are used for parts common to the pseudo force sense generator 12 of the pseudo force sense generator 12 ″.

[Modification 3 of the first and second embodiments]
The pseudo force generation device may include a plurality of motion members. For example, FIG. 10 shows a mechanical property model Md of the pseudo force sense generator including two or more moving members and a mechanical property model Ms of the skin of the finger in contact with the grip part.

Mechanical properties model Md In this example, the mass point _{M 1} of the mass _{m 1,} the respective mass _{m j} (however, j = 2, ..., a j_max, j_max is an integer of 3 or more) mass _{M j,} mass of Dynamics composed of a spring having an elastic coefficient k _j connecting M ₁ and a mass point M _j , a damper having a viscosity coefficient b _j , and a periodic Lorentz force f _j acting on the mass point M ₁ and the mass point M _j according to the drive voltage Vout _j Represents the characteristics of the system. The position of the mass point M ₁ with respect to the reference origin O ₁ is expressed as x ₁ . The position of the mass point M _j with respect to the reference origin O _j is expressed as x _j . However, the reference origin O _j is a point where the relative position with respect to the center of gravity of the finger is fixed. Also, x _j is positive in the right direction in FIG. 3A from the center of gravity of the finger in contact with the grip portion of the pseudo force sense generating device, and negative in the opposite direction. x _j time derivative, ie speed

Is written. Due to the limitation of the description notation, this may be expressed as x ^· _j in this specification. x ^· _j time derivative, ie acceleration

Is written. In this specification, this may be expressed as x ^·· _j .

The skin dynamic characteristic model Ms in this example represents characteristics of a dynamic system including a spring having an elastic coefficient k ₁ and a damper having a viscosity coefficient b ₁ between the mass point M ₁ and the center of gravity of the finger in contact with the grasping portion. Here, the force applied to the skin of the finger that contacts the gripping part is expressed as fs.

If the motions of the mass points M ₁ and M _j are constrained by the motion in one axis direction, the following equation of motion holds.

However, f ₁ represents the force applied from the skin of the finger to the mass point M _1. The inverse dynamics model Mc to be controlled may be set using these equations (14) to (17).

[Modification 4 of the first and second embodiments]
When the strength of the force for grasping the pseudo force sense generating device 12 changes, the skin dynamic characteristic model changes, so that the inverse dynamics model Mc to be controlled also changes. In order to cope with such a change, for example, a sensor for measuring the force [N] applied to the skin is installed in the grip portion 126 of the pseudo force generation device 12, and the force detected by the sensor at a certain time and the detection time If the absolute value of the error from the force in the target response waveform pattern corresponding to is greater than or equal to a predetermined threshold value, it is determined that the inverse dynamics model Mc to be controlled is not appropriate, and the parameter of the skin dynamic characteristic model Ms is set to a predetermined value. The inverse dynamics model Mc to be controlled may be recalculated by changing the method.

  The process for determining that the inverse dynamics model Mc to be controlled is not appropriate is not limited to this, and the error between the force time-series data detected by the sensor and the force in the target response waveform pattern corresponding to the time when the data was acquired Other methods such as recalculating the inverse dynamics model Mc to be controlled may be used when the sum of squares is equal to or greater than a predetermined threshold.

  Regarding the method of changing the parameters of the skin mechanical property model Ms, for example, the skin mechanical properties of the finger 100 according to the force applied to the finger 100 when the pseudo force sense generating device 12 is gripped by the finger 100. A combination of the elastic coefficient k1 of the spring constituting the model Ms and the viscosity coefficient b1 of the damper is stored in a plurality of storage units, and the elastic coefficient k1 and the viscosity coefficient b1 of the skin mechanical property model Ms of the finger 100 according to the error. The value may be changed. For example, a plurality of combinations of the elastic coefficient k1 and the viscosity coefficient b1 are stored in the storage unit in ascending order or descending order of the strength of gripping the pseudo force generation device 12, and the force detected by the sensor at a certain time is detected. One above or below the combination of the elastic coefficient k1 and the viscosity coefficient b1 being used in the current inverse dynamics model Mc depending on whether the force is smaller or larger than the force in the target response waveform corresponding to the time. It is also possible to select the next combination.

  It is assumed that the pseudo force sense generating device 12 includes an acceleration sensor instead of a force sensor, and an estimated value of acceleration generated in the pseudo force sense generating device 12 estimated from the target response waveform pattern and an acceleration measured by the acceleration sensor It is also possible to determine whether or not it is necessary to recalculate the inverse dynamics model Mc to be controlled using the above error.

[Third Embodiment]
In this embodiment, contrary to the above-described embodiment, the control signal waveform is applied to the forward dynamics model to be controlled, and the response waveform is estimated.

<Configuration>
As illustrated in FIG. 11, the waveform estimation device 31 of this embodiment includes a storage unit 311, an input unit 312, a model application unit 313, and an output unit 314. The waveform estimation device 31 is a device configured by, for example, the above-described general-purpose or dedicated computer executing a predetermined program.

<Pretreatment>
In the preprocessing of the present embodiment, the forward dynamics model Mp to be controlled is set and stored in the storage unit 311. The forward dynamics model Mp to be controlled according to the present embodiment has the position, speed, and acceleration of the gripper 126 of the pseudo force sense generator 12 with respect to the control signal waveform (control signal waveform) for driving the moving member 125 described above. , And a mathematical model for obtaining a response waveform corresponding to at least one of the forces applied to the skin in contact with the grip portion 126. The forward dynamics model Mp to be controlled according to the present embodiment is obtained based on the dynamic characteristic model Md of the pseudo force sense generator 12 and the dynamic characteristic model Ms of the skin in contact with the grasping portion 126. Examples of the mechanical property model Md and the mechanical property model Ms are as described above. The inverse function or approximate inverse function of the forward dynamic model Mp to be controlled is the aforementioned inverse dynamic model Mc to be controlled. Examples of the forward dynamics model Mp to be controlled are the relational expressions (3A) to (10A) described above.

<Response waveform estimation process>
In the response waveform estimation processing of this embodiment, the control signal waveform is input to the input unit 312. The control signal waveform is a control signal waveform for driving the moving member 125 described above. The control signal waveform is a time waveform of a physical quantity necessary for controlling the moving member 125 (a waveform made up of physical quantities at each time). The control signal waveform is, for example, a periodic waveform in which a positive waveform and a negative waveform are asymmetric (for example, FIGS. 4A, 5A, 6A, and 7A). In addition, which physical quantity time waveform is used as the control signal waveform depends on the configuration of the forward dynamics model Mp to be controlled. That is, the control target forward dynamics model Mp is a model for obtaining a response waveform from the control signal waveform, and the control signal waveform input to the input unit 112 is a time waveform corresponding to the control target forward dynamics model Mp. For example, when the forward dynamics model Mp to be controlled is the relational expression (3A), the time waveform of the current i ₂ is set as the control signal waveform. For example, when the forward dynamics model Mp to be controlled is the relational expression (4A), the time waveform of the voltage Vout is set as the control signal waveform.

The control signal waveform is sent to the model application unit 313. The model application unit 313 applies the control signal waveform to the forward dynamics model Mp to be controlled read from the storage unit 111, and controls the asymmetric motion of the pseudo force generation device 12 with the control signal waveform. A response waveform corresponding to at least one of speed, acceleration, and force applied to the skin is estimated. That is, the model application unit 313 substitutes the physical quantity of each time of the control signal waveform into the forward dynamics model Mp to be controlled, obtains a function value corresponding thereto, and outputs the time waveform of the function value as a “response waveform”. . For example, when the forward dynamics model Mp to be controlled is the relational expression (3A), the model application unit 313 calculates fs = F ₂ (i ₂ ) at each time from the current i _{2 at} each time of the control signal waveform. Output as response waveform. For example, when the forward dynamics model Mp to be controlled is the relational expression (4A), the model applying unit 313 calculates fs = F _R (Vout) at each time from the voltage Vout at each time of the control signal waveform, and a response waveform. Output as. The obtained response waveform is sent to the output unit 314, and the output unit 314 outputs this response waveform.

<Features of this embodiment>
According to the present embodiment, it is possible to estimate the control signal waveform and the response waveform according to the configuration and way of holding the pseudo force sense generator 12.

[Fourth Embodiment]
The fourth embodiment is a modification of the third embodiment, and the forward dynamics model Mp to be controlled is the dynamic characteristic model Md of the simulated force sense generator 12, the dynamic characteristic model Ms of the skin, and the control signal waveform and the electric circuit. The difference is based on the electric circuit characteristic model Me representing the relationship with the output of.

<Configuration>
As illustrated in FIG. 11, the waveform estimation device 41 of this embodiment includes a storage unit 411, an input unit 312, a model application unit 413, and an output unit 314. The waveform estimation device 41 is a device configured by, for example, the aforementioned general-purpose or dedicated computer executing a predetermined program.

In the preprocessing of the present embodiment, the forward dynamics model Mp to be controlled is set and stored in the storage unit 411. The forward dynamics model Mp to be controlled according to the present embodiment has the position, speed, and acceleration of the gripper 126 of the pseudo force sense generator 12 with respect to the control signal waveform (control signal waveform) for driving the moving member 125 described above. , And a mathematical model for obtaining a response waveform corresponding to at least one of the forces applied to the skin in contact with the grip portion 126. The forward dynamics model Mp to be controlled in this embodiment is obtained based on the above-described mechanical characteristic model Md, dynamic characteristic model Ms, and electric circuit characteristic model Me. For example, a relational expression based on any one of the relational expressions (3A) to (10A) and the expression (11) is set as the forward dynamics model Mp to be controlled. For example, the relational expression fs = F _R (H (S) · Vin) = F _RH (Vin) (18) based on the relational expression (4A) and the expression (11).
May be the forward dynamics model Mp to be controlled. Or the relational expression x ^·· ₁ = F _m (H (S) · Vin / R) = F _mRH (Vin) (19) based on the relational expression (5A), the expression (11), and i ₂ = Vout / R
May be the forward dynamics model Mp to be controlled.

<Response waveform estimation process>
In the response waveform estimation processing of this embodiment, the control signal waveform is input to the input unit 312. The control signal waveform is a control signal waveform for driving the moving member 125 described above. As described above, which physical quantity time waveform is used as the control signal waveform depends on the configuration of the forward dynamics model Mp to be controlled. For example, when the forward dynamics model Mp to be controlled is the relational expression (18), the time waveform of the voltage Vin is set as the control signal waveform.

The control signal waveform is sent to the model application unit 413. The model application unit 413 applies the control signal waveform to the forward dynamics model Mp to be controlled read from the storage unit 111, and controls the asymmetric motion of the pseudo force generation device 12 with the control signal waveform. A response waveform corresponding to at least one of speed, acceleration, and force applied to the skin is estimated. For example, when the forward dynamics model Mp to be controlled is the relational expression (18), the model applying unit 313 calculates fs = F _RH (Vin) at each time from the voltage Vin at each time of the control signal waveform, and a response waveform. Output as.

<Features of this embodiment>
According to the present embodiment, it is possible to estimate the response waveform according to the control signal waveform, the configuration and holding method of the pseudo force sense generator 12, and the electric circuit 13.

[Fifth Embodiment]
An estimated waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the skin, obtained by applying a trial signal waveform as a control signal waveform to the forward dynamics model Mp to be controlled, is input A target for controlling the asymmetric motion of the pseudo force sense generator 12 by adjusting the trial signal waveform based on the error from the target response waveform pattern (for example, using the steepest descent method so that the error is minimized). A control signal waveform may be obtained.

<Configuration>
As illustrated in FIG. 12, the system of this embodiment includes a control signal estimation device 51 that obtains a target control signal waveform from a target response waveform pattern, a pseudo force generation device 12, and an electric circuit 13. The control signal estimation device 51 includes a storage unit 311, an input unit 112, a model application unit 313, a trial signal update unit 514, and an error calculation unit 515. The control signal estimation device 51 is a device configured by, for example, the above-described general-purpose or dedicated computer executing a predetermined program.

<Pretreatment>
As described in the third embodiment, the forward dynamics model Mp to be controlled is set and stored in the storage unit 311. The forward dynamics model Mp to be controlled in this embodiment is in contact with the position, speed, acceleration, and gripping portion 126 of the gripping portion 126 of the pseudo force sense generator 12 with respect to the control signal waveform for driving the motion member 125. It is a mathematical model for obtaining a response waveform corresponding to at least one of the forces applied to the skin. The forward dynamics model Mp to be controlled in this embodiment is based on the dynamic characteristic models Md and Ms.

<Target control signal waveform estimation process>
In the target control signal waveform estimation process of this embodiment, the target response waveform pattern TA (t) is input to the input unit 112. t represents time. The target response waveform pattern TA (t) corresponds to at least one of the position, speed, acceleration, and force applied to the skin of the finger 100 as a target. Which physical quantity time waveform pattern is the target response waveform pattern TA (t) depends on the configuration of the forward dynamics model Mp to be controlled. That is, the forward dynamics model Mp to be controlled is a model for obtaining a response waveform from the control signal waveform, and the target response waveform pattern TA (t) input to the input unit 112 is the response waveform of the forward dynamics model Mp to be controlled. It corresponds. For example, when the forward dynamics model Mp to be controlled is the relational expression (3A), the time waveform of the target force fs is set as the target response waveform pattern TA (t). The target response waveform pattern TA (t) is sent to the error calculator 515 (step S51).

  The trial signal update unit 514 sets a trial signal waveform TR (t) as a control signal waveform of the forward dynamics model Mp to be controlled. The initial value of the trial signal waveform TR (t) is predetermined, for example, TR (t) = 0. The trial signal waveform TR (t) is sent to the model application unit 313 (step S52).

The model application unit 313 applies the transmitted trial signal waveform TR (t) as a control signal waveform to the forward dynamics model Mp to be controlled, and at least one of the position, velocity, acceleration, and force applied to the skin of the gripping unit. An estimated waveform W (t) corresponding to is obtained. For example, when the forward dynamics model Mp to be controlled is the relational expression (3A), W (t) = F ₂ (TR (t)) is calculated. The estimated waveform W (t) is sent to the error calculator 515 (step S53).

  The error calculation unit 515 determines whether a predetermined end condition is satisfied based on the error δ (t) between the target response waveform pattern TA (t) and the estimated waveform W (t). Examples of termination conditions are that error δ (t) for all times t is less than or equal to a threshold, that the sum of squares of error δ (t) in a predetermined time interval (for example, one period) is less than or equal to the threshold, trial For example, the error δ (t) corresponding to the update of the signal waveform TR (t) or the change amount of the sum of squares is equal to or less than a threshold value. If it is determined that the predetermined end condition is satisfied, the error calculation unit 515 sends information indicating that the end condition is satisfied to the trial signal update unit 514. On the other hand, when it is determined that the predetermined end condition is not satisfied, the error calculation unit 515 sends information indicating that the end condition is not satisfied to the trial signal update unit 514 (step S54).

  When information indicating that the termination condition is not satisfied is sent to the trial signal update unit 514, the trial signal update unit 514 sets a new trial signal waveform TR (t) (for example, the error δ (t) is The data is sent to the model application unit 313 (so as to be smaller) (step S55). In this case, steps S53 and S54 are executed again. On the other hand, when information indicating that the end condition is satisfied is sent to the trial signal update unit 514, the trial signal update unit 514 outputs the latest trial signal waveform TR (t) as the target control signal waveform (step S56). .

  A control signal (control voltage or control current) corresponding to the obtained target control signal waveform is applied to the electric circuit 13. The electric circuit 13 sends a corresponding drive signal to the pseudo force sense generator 12. The simulated force sense generator 12 is driven based on this, and presents a simulated force sense to the finger 100 gripping the grip portion 126.

<Features of this embodiment>
In this embodiment, the target control signal waveform for controlling the pseudo force sense generator 12 can be estimated from the target response waveform pattern for presenting the pseudo force sense. Thereby, even when the configuration and the way of holding the pseudo force sense generator 12 are changed, an appropriate target control signal waveform can be estimated according to each. In addition, unlike the case where the inverse dynamics model to be controlled is obtained by an inverse function or approximate inverse function, the target control signal waveform can be accurately estimated even if the forward dynamics model to be controlled is nonlinear.

  The mechanical characteristic models Md and Ms are obtained by modeling the state in which the grip portion 126 (that is, the grip portion of the actual machine) of the pseudo force sense device 12 controlled based on the target control signal waveform is gripped. Alternatively, it may be a model of a state in which the grip portion of another device that is the same as or similar to the simulated force sense generator 12 is gripped. In the former case, a more accurate target control signal waveform can be estimated.

[Sixth Embodiment]
The sixth embodiment is a modification of the fifth embodiment, and the forward dynamics model Mp to be controlled is the dynamic characteristic model Md of the pseudo force sense generator 12, the dynamic characteristic model Ms of the skin, and the control signal waveform and the electric circuit. The difference is based on the electric circuit characteristic model Me representing the relationship with the output of.

  As illustrated in FIG. 12, the system of this embodiment includes a control signal estimation device 61, a pseudo force generation device 12, and an electric circuit 13. The control signal estimation device 61 includes a storage unit 411, an input unit 112, a model application unit 413, a trial signal update unit 514, and an error calculation unit 515. The control signal estimation device 61 is a device configured by, for example, the above-described general-purpose or dedicated computer executing a predetermined program.

<Pretreatment>
As described in the fourth embodiment, the forward dynamics model Mp to be controlled is set and stored in the storage unit 311. The forward dynamics model Mp to be controlled in this embodiment is in contact with the position, speed, acceleration, and gripping portion 126 of the gripping portion 126 of the pseudo force sense generator 12 with respect to the control signal waveform for driving the motion member 125. It is a mathematical model for obtaining a response waveform corresponding to at least one of the forces applied to the skin. The forward dynamics model Mp to be controlled in this embodiment is based on the mechanical characteristic models Md and Ms and the electric circuit characteristic model Me.

<Target control signal waveform estimation process>
The process is the same as the target control signal waveform estimation process of the fifth embodiment except that the forward dynamics model Mp to be controlled based on the mechanical characteristic models Md and Ms and the electric circuit characteristic model Me is used.

<Experimental result>
The comparison experiment result of this form is illustrated.
<< Experimental result 2: [T1, T2] = [2,10] msec >>
13A to 13C show experimental results according to this embodiment. In this experiment, the control signal waveform of FIG. 13A was obtained using the target response waveform pattern indicated by the broken line of FIG. 13B. This target response waveform pattern (FIG. 13B) is a time waveform of the force [N] applied to the skin of the finger 100 that is in contact with the grip portion 126. This target response waveform pattern is a periodic pattern in which a time interval T1 of 2 [msec] and a subsequent time interval T2 of 10 [msec] are repeated, and the time interval T1 and the time interval T2 are 1 in one cycle. One period is 12 [msec] consisting of a time interval T1 and a time interval T2. The response waveform of the force [N] applied to the skin of the finger 100 in contact with the grasping portion 126 by the simulated force sense generator 12 controlled by the target response waveform pattern (FIG. 13A) is shown by a solid line in FIG. 13B. FIG. 13B shows that a response waveform that approximates the target response waveform pattern is obtained. FIG. 13C shows the waveform of the position [m] of the grip portion 12 at this time.

<< Experimental result 3: [T1, T2] = [3, 16] msec >>
14A to 14C show experimental results according to this embodiment. Unlike the experimental result 4, [T1, T2] = [3, 16] msec. Other conditions are the same as those in Comparative Experiment Result 1.

  4B and FIG. 13B, FIG. 4C and FIG. 13C, FIG. 6B and FIG. 14B, and FIG. 6C and FIG. 14C. It can be seen that there is little fine vibration of the waveform, while the difference between the maximum value in the positive direction and the maximum value in the negative direction is large, and the asymmetry is large. Thereby, it turns out that a clear force sense can be shown in this form rather than the conventional method.

<Features of this embodiment>
Also in this embodiment, the target control signal waveform for controlling the pseudo force sense generator 12 can be estimated from the target response waveform pattern for presenting the pseudo force sense. Thereby, even when the configuration and the way of holding the pseudo force sense generator 12 are changed, an appropriate target control signal waveform can be estimated according to each. Also, unlike the case where the inverse dynamics model of the controlled object is obtained by an inverse function or approximate inverse function, even if the forward dynamics model of the controlled object is nonlinear (for example, when the output of the electric circuit is saturated), its nonlinear characteristics As a result, the target control signal waveform can be accurately estimated.

  The mechanical characteristic models Md and Ms are obtained by modeling the state in which the grip portion 126 (that is, the grip portion of the actual machine) of the pseudo force sense device 12 controlled based on the target control signal waveform is gripped. Alternatively, it may be a model of a state in which the grip portion of another device that is the same as or similar to the simulated force sense generator 12 is gripped. Similarly, the electric circuit characteristic model Me may be a model of the electric circuit 13 (that is, an actual machine) to which a control signal having a target control signal waveform is given, or other electric circuits that are the same as or similar to the electric circuit 13. It may be a model of a circuit. In either case, a more accurate target control signal waveform can be estimated in the former case.

[Modifications of Fifth and Sixth Embodiments]
As a modified example of the fifth and sixth embodiments, even if the simulated force sense generators 12 ′ and 12 ″ illustrated in the modified examples 1 and 2 of the first and second embodiments are used instead of the simulated force sense generator 12. As exemplified in Modification 3 of the first and second embodiments, the simulated force sense generator may include a plurality of motion members.

[Seventh Embodiment]
Using the target control signal waveform of any one of the first to sixth embodiments or its modification, the base mechanism supported directly or indirectly on the first skin, and depending on the target control signal waveform with respect to the base mechanism Alternatively, a pseudo force sense generating device having a contact mechanism that performs a periodic asymmetric motion and applies a force based on the asymmetric motion to the second skin directly or indirectly in contact may be driven. However, the mass of this contact mechanism is smaller than the mass of the base mechanism.

<Configuration>
As illustrated in FIGS. 15A to 15D, the pseudo force generation apparatus 1000 according to the present embodiment includes a base member 1001, a support unit 1026-1, and a motion member that performs asymmetric vibration (periodic asymmetric motion) with respect to the support unit 1026-1. It has a vibrator 1020-1 including 1025-1 and a contact portion 1003-1. The moving member 1025-1 of this embodiment performs asymmetric vibration along the D-1 axis while being supported by the support 1022-1 based on the drive signal output from the electric circuit 13 described above. The base portion 1001 is directly or indirectly supported by the skin 201 (first skin) of the hand 200, and the contact portion 1003-1 is attached to the skin 202 (second skin) of the hand 200 that is in direct or indirect contact. Gives a force based on asymmetric vibration. This asymmetric vibration is a vibration for causing a pseudo force sense to be perceived.

In this embodiment, a mechanism including the base portion 1001 and the support portion 1026-1 (for example, a mechanism including the base portion 1001 and the support portion 1026-1) corresponds to the “base mechanism”. A mechanism including the contact portion 1003-1 and the moving member 1025-1 (for example, a mechanism including the contact portion 1003-1, the moving member 1025-1, and a member connecting them) corresponds to the “contact mechanism”. The mass m ₁ of the “contact mechanism” is smaller than the mass m ₂ of the “base mechanism”. For example, the total mass of the mechanism including the contact portion 1003-1, the motion member 1025-1, and the member that connects them is smaller than the total mass of the base portion 1001 and the support portion 1026-1. Preferably, the mass of the “contact mechanism” is greater than zero and less than or equal to one third of the mass of the “base mechanism”. For example, the sum total of the mass of the mechanism including the contact portion 1003-1, the motion member 1025-1, and the member that connects them is one third or less of the total mass of the base portion 1001 and the support portion 1026-1. In such a configuration, even when the mass m ₁ + m _{2 of the} entire system is large, since the mass m ₁ of the “contact mechanism” is small, a sufficiently large force is transmitted from the contact portion 1003-1 to the skin 202. . As a result, the vibrator 1020-i having the same stroke and output can give a larger deformation to the skin 202 than the conventional method. Further, since the relative displacement between the moving member 1025-i and the support portion 1026-i can be reduced, the vibrator 1020-i having a smaller stroke can be used. By using such a mechanism to vibrate the vibrator 1020-i asymmetrically, a pseudo force sense such as a pulling sensation can be efficiently perceived.

<Base 1001>
As illustrated in FIGS. 15A to 15D, the base portion 1001 of this embodiment is a plate-like member, and a concave portion 1001d-1 in which the vibrator 1020-1 is disposed is provided on the upper surface 1001b side. The base 1001 may be anything. Examples of the base 1001 are mobile terminal devices such as a smart phone terminal device, a tablet terminal device, an electronic book reader device, a mobile phone terminal device, a notebook personal computer, and a portable game machine. The controller 100 and other electronic devices may be the base 1001, and members other than the electronic devices may be the base 1001. The base 1001 may include a mobile terminal device such as a mobile phone terminal device and other members.

<Oscillator 1020-1>
A support portion 1026-1 of the vibrator 1020-1 is attached to the bottom surface 1001e-1 of the recess 1001d-1. Note that “α is attached to β” means that α is fixed to β, α is connected to β, α is detachably held on β, or α Is held in β with “play (margin)” or “backlash”. Further, “α is attached to β” is a concept including not only that α is directly attached to β but also that α is indirectly attached to β through an inclusion. . “Α is attached to β” is also synonymous. Accordingly, the vibrator 1020-1 is supported by the base 1001, and the vibrator 1020-1 is disposed inside the recess 1001d-1. The motion member 1025-1 of the vibrator 1020-1 is capable of performing asymmetric vibration with respect to the support 1022-1 along the D-1 axis while being supported by the support 1026-1. The D-1 axis in this embodiment is an axis along the short direction of the base 1001, and the motion member 1025-1 is asymmetrically vibrated in the short direction of the base 1001. Hereinafter, a specific configuration of the vibrator 1020-1 will be exemplified.

  As illustrated in FIGS. 16A and 16B, the vibrator 1020-i (i is a positive integer) includes, for example, a support portion 1026-i including a case 1027-i and a guide 1021-i, and springs 1022-i and 1023-. This is a linear actuator having i (elastic body), a coil 1024-i, a moving member 1025-i that is a permanent magnet, and connecting portions 1029c-i, 1029d-i, 1029a-i, and 1029b-i. Both the case 1027-i and the guide 1021-i in this embodiment are hollow members having a shape in which a part of both open ends of a cylinder (for example, a cylinder or a polygonal cylinder) is closed. However, the guide 1021-i is smaller than the case 1027-i and has a size that can be accommodated in the case 1027-i. Case 1027-i, guide 1021-i, and connection part 1029a-i, 1029b-i, 1029c-i, 1029d-i are comprised from synthetic resins, such as an ABS resin, for example. The springs 1022-i and 1023-i are, for example, a helical spring or a leaf spring made of metal or the like. The springs 1022-i and 1023-i preferably have the same elastic coefficient (spring constant), but may be different from each other. The moving member 1025-i is, for example, a cylindrical permanent magnet, and one end portion 1025 a-i side in the longitudinal direction is an N pole, and the other end portion 1025 b-i side is an S pole. The coil 1024-i is, for example, a continuous enameled wire, and includes a first winding portion 1024a-i and a second winding portion 1024b-i.

  The moving member 1025-i is accommodated in the guide 1021-i and supported so as to be slidable in the longitudinal direction. Although details of such a support mechanism are not shown, for example, a straight rail along the longitudinal direction is provided on the inner wall surface of the guide 1021-i, and this rail is slidably supported on the side surface of the motion member 1025-i. A rail support is provided. One end of a spring 1022-i is fixed to the inner wall surface 1021a-i on one end side in the longitudinal direction of the guide 1021-i (that is, one end of the spring 1022-i is supported by the guide 1021-i), and the spring 1022- The other end of i is fixed to the end portion 1025a-i of the motion member 1025-i (that is, the end portion 1025a-i of the motion member 1025-i is supported by the other end of the spring 1022-i). Further, one end of the spring 1023-i is fixed to the inner wall surface 1021b-i on the other end side in the longitudinal direction of the guide 1021-i (that is, one end of the spring 1023-i is supported by the guide 1021-i), The other end of the spring 1023-i is fixed to the end portion 1025b-i of the motion member 1025-i (that is, the end portion 1025b-i of the motion member 1025-i is supported by the other end of the spring 1023-i. )

A coil 1024-i is wound around the outer periphery of the guide 1021-i. However, the end 1025a-i side of the moving member 1025-i (N pole side), and the first winding portion 1024a-i is wound _{A 1} direction (direction toward the back to the front), the end portion in 1025b-i side (S-pole side), a second winding portion 1024b-i is wound around the _{a 1} direction in the opposite direction to the direction of _{B 1} direction (direction toward the front to the back). That is, when viewed from the end portion 1025a-i side (N pole side) of the moving member 1025-i, the first winding portion 1024a-i is wound clockwise, and the second winding portion 1024b-i is counterclockwise. It is wrapped around Further, in a state where the motion member 1025-i is stopped and the elastic force from the springs 1022-i and 1023-i is balanced, the end portion 1025 a-i side (N pole side) of the motion member 1025-i is the first winding. It is desirable that the end portion 1024a-i be disposed in the region of the second winding portion 1024b-i.

  The guide 1021-i, the springs 1022-i, 1023-i, the coil 1024-i, and the motion member 1025-i arranged and configured as described above are accommodated in the case 1027-i, and the guide 1021-i is the case. It is fixed inside 1027-i. That is, the relative position of the case 1027-i with respect to the guide 1021-i is fixed. However, the longitudinal direction of the case 1027-i coincides with the longitudinal direction of the guide 1021-i and the longitudinal direction of the moving member 1025-i.

  A through hole 1028a-i is provided on the inner wall surface 1021a-i side of the case 1027-i and the guide 1021-i, and a through hole 1028b-i is provided on the inner wall surface 1021b-i side. A rod-like connecting portion 1029a-i is inserted into the through-hole 1028a-i, and a rod-like connecting portion 1029b-i is inserted into the through-hole 1028b-i. One end side of the connecting portion 1029a-i contacts the end portion 1025a-i side of the motion member 1025-i, and the other end side of the connecting portion 1029a-i is a connecting portion 1029c-i disposed outside the case 1027-i. It is supported on one end side. One end side of the connecting portion 1029b-i contacts the end portion 1025b-i side of the motion member 1025-i, and the other end side of the connecting portion 1029b-i is a connecting portion 1029d-i arranged outside the case 1027-i. It is supported on one end side. One end side of the connecting portion 1029a-i may be connected to the end portion 1025a-i side of the motion member 1025-i, or may not be connected. One end side of the connecting portion 1029b-i may or may not be connected to the end portion 1025b-i side of the motion member 1025-i. For example, the end portions 1025a-i and 1025b-i of the motion member 1025-i may be sandwiched between one end side of the connecting portion 1029a-i and one end side of the connecting portion 1029d-i. However, it is necessary that the connecting portions 1029c-i, 1029d-i, 1029a-i, and 1029b-i move with the movement of the moving member 1025-i. That is, the connecting portions 1029a-i, 1029b-i, 1029c-i, and 1029d-i must move together with the moving member 1025-i. In addition, one end side of the connecting portion 1029a-i may be integrated with the end portion 1025a-i side of the motion member 1025-i, and one end side of the connecting portion 1029b-i is the end portion 1025b-i of the motion member 1025-i. It may be integral with the side.

The coil 1024-i applies a force corresponding to the applied current to the motion member 1025-i, whereby the motion member 1025-i is periodically asymmetrically moved (the guide 1021-i) with respect to the guide 1021-i. Periodic translational reciprocating motion with asymmetry in the axial direction with respect to. That is, when an electric current is applied to the coil 1024-i to _{A 1} direction _{(B 1} direction), by reaction of the Lorentz force is described in Fleming's left-hand rule, _{C 1} direction moving member 1025-i (movement member 1025- A force in the direction from the north pole of i to the south pole (right direction) is applied (FIG. 16A). Conversely, _{A 2} direction in the coil 1024-i when _{(B 2} direction) electric current, _{C 2} direction movement member 1025-i (the direction toward the S pole of the moving member 1025-i to the N pole: left) Is applied (Figure 16B). However, _{A 2} direction is opposite the direction of _{A 1} direction. By these operations, kinetic energy is given to the system including the moving member 1025-i and the springs 1022-i and 1023-i. As a result, the position and acceleration of the moving member 1025-i with respect to the case 1027-i (the position and acceleration in the axial direction with reference to the guide 1021-i) are changed, and accordingly, the connecting portions 1029 c-i and 1029 d. The position and acceleration of -i, 1029a-i, and 1029b-i can also be changed. In other words, the motion member 1025-i is supported by the support portion 1026-i based on the given drive signal (drive voltage Vout or drive current), and is applied to the support portion 1026-i along the Di axis. Asymmetric vibration is performed, and the connecting portions 1029c-i, 1029d-i, 1029a-i, and 1029b-i also perform asymmetric vibration along the Di axis.

Note that the configuration of the vibrator 1020-i is not limited to that shown in FIGS. 16A and 16B. For example, the first winding portion 1024a-i of the coil 1024-i to the end 1025a-i side of the moving member 1025-i are wound _{A 1} direction, the coil 1024-i is wrapped around the end 1025b-i side The structure which is not done may be sufficient. Conversely, the second winding portion 1024b-i are wound around the _{B 1} direction, the coil 1024-i to the end 1025a-i side of the moving member 1025-i of the coil 1024-i to the end 1025b-i side The structure which is not wound may be sufficient. Alternatively, the first winding portion 1024a-i and the second winding portion 1024b-i may be different coils. That is, the 1st winding part 1024a-i and 2nd winding part 1024b-i may not be electrically connected, and the structure to which a mutually different electric signal is given may be sufficient.

<Contact part 1003-1>
A contact part 1003-1 is attached to the moving member 1025-1 of the vibrator 1020-1, and the contact part 1003-1 is thereby supported by the vibrator 1020-1. That is, the contact portion 1003-1 is attached to the motion member 1025-i and is capable of vibrating with respect to the support portion 1026-1. As illustrated in FIGS. 15A to 15D, the contact portion 1003-1 of this embodiment has a plate-like member whose planar shape (the shape of the plate surfaces 1003a-1 and 1003b-1) is substantially rectangular (for example, a rectangle or a square). It is. The sizes of the plate surfaces 1003a-1 and 1003b-1 are smaller than the area of the opening of the recess 1001d-1 of the base 1001. That is, the area of the plate surface 1003 a-1 of the contact part 1003-1 is narrower than the area of the upper surface 1001 b of the base part 1001. It is desirable that the contact portion 1003-1 has a hardness capable of conducting vibration based on the asymmetric vibration of the motion member 1025-i and is made of a material that is as light as possible. An example of such a material is a synthetic resin such as an ABS resin.

  The contact portion 1003-1 is disposed on the opening side of the recess 1001d-1 so that the plate surfaces 1003a-1 and 1003b-1 are substantially parallel to the upper surface 1001b. An example of “substantially parallel” is parallel. The plate surface 1003a-1 is directed outward from the base portion 1001, and is disposed on substantially the same plane as the upper surface 1001b. The plate surface 1003b-1 is directed to the bottom surface 1001e-1 side of the recess 1001d-1, and the other end sides of the connecting portions 1029c-i and 1029d-i are attached to the plate surface 1003b-1. That is, the moving member 1025-1 of the vibrator 1020-1 is attached to the contact portion 1003-1 via the above-described connecting portions 1029c-i, 1029d-i, 1029a-i, and 1029b-i. Thereby, the contact part 1003-1 and the connection parts 1029c-i, 1029d-i, 1029a-i, and 1029b-i are supported by the base part 1001 and the support part 1026-1. There is a gap between the contact portion 1003-1 and the inner wall surface 1001f-1 of the recess 1001d-1, and these do not contact each other. Alternatively, the contact portion 1003-1 may be supported in a direction substantially orthogonal to the moving direction of the vibrator 1020-1 by a rail or the like (that is, the contact portion 1003-1 may be supported by the vibrator 1020-1 by a rail or the like). May be supported so as not to move in a direction substantially orthogonal to the direction of motion of Thereby, the contact part 1003-1 can vibrate in the direction along the upper surface 1001b of the base part 1001 with respect to the base part 1001 and the support part 1026-i. That is, the contact part 1003-1 can vibrate in the vibration direction of the vibrator 1020-1 substantially parallel to the upper surface 1001b of the base part 1001 with respect to the base part 1001 and the support part 1026-i.

<Mass of system>
As described above, the “base mechanism” of the present embodiment is a mechanism including the base portion 1001 and the support portion 1026-1, for example, a mechanism including the base portion 1001 and the support portion 1026-1. The “base mechanism” may further include at least one of springs 1022-i, 1023-i, and coils 1024-i. The “contact mechanism” is a mechanism including the contact portion 1003-1 and the motion member 1025-1. For example, the contact portion 1003-1, the motion member 1025-1, and the connection portions 1029a-i, 1029b-i, and 1029c- i, 1029d-i. The mass m ₁ of the “contact mechanism” is smaller than the mass m ₂ of the “base mechanism”. This effectively presents a pseudo force sense. Preferably, the mass m ₁ of the “contact mechanism” is greater than zero and not more than one third of the mass of the mass m ₂ of the “base mechanism”. In other words, 0 <m ₁ / m ₂ ≦ 1/3. Thereby, a pseudo force sense can be presented more efficiently.

<Operation>
As illustrated in FIG. 15D, when the pseudo force generation device 1000 is used, the bottom surface 1001c side of the base 1001 is held (gripped) by the skin 201 such as the palm of the hand 200 or a finger other than the thumb, and the contact unit 1003-1 Skin 202 such as the thumb of hand 200 contacts. The base 1001 may be in direct contact with the skin 201 or may be held via an inclusion. Similarly, the contact portion 1003-1 may be in direct contact with the skin 202 or may be in contact with inclusions. However, a force based on the vibration of the contact portion 1003-1 must be transmitted to the skin 202.

As described above, a control signal (control voltage or control current) corresponding to the target control signal waveform of any one of the first to sixth embodiments or its modification is applied to the electric circuit 13. The electric circuit 13 sends a corresponding drive signal (drive voltage Vout or drive current) to the vibrator 1020-1 of the pseudo force sense generator 1000. However, in the dynamic characteristic model Md of the pseudo force generation apparatus 1000 of the present embodiment, the “contact mechanism” corresponds to the mass point M ₁ and the “base mechanism” corresponds to the mass point M ₂ . In both x ₁ and x ₂ , the right direction in FIG. 15D from the center of gravity of the hand 200 is positive, and the left direction in FIG. 15D from the center of gravity of the hand 200 is negative. The mechanical property model Ms of this embodiment is a mechanical property model of the skin 202. Note that the mass m ₂ of the “base mechanism” is larger than the mass m ₁ of the “contact mechanism”. Therefore, the magnitude of the force applied to the skin 201 as the vibrator 1020-1 is driven is smaller than the magnitude of the force applied to the skin 202. In this embodiment, the mechanical characteristic model of the skin 201 is omitted.

This driving signal, a desired direction (FIGS. 16A and 16B: _{C 1} direction or _{C 2} direction) motion member 1025-1 and the period T1 to flow direction of the current in the coil 1024-1 giving acceleration, the other Period T2 is periodically repeated. At that time, the ratio (reversal ratio) between the period (time) in which current flows in a predetermined direction and the other period (time) is biased to one of the periods. In other words, a periodic current having a ratio of the period T1 occupying one period different from the ratio of the period T2 occupying the period is supplied to the coil 1024-1. Thereby, the moving member 1025-1 asymmetrically vibrates along the D-1 axis with respect to the support portion 1026-1. The asymmetric vibration caused by the moving member 1025-1 is transmitted to the contact part 1003-1 via the connecting parts 1029a-1, 1029b-1, 1029c-1, and 1029d-1, and the contact part 1003-1 is asymmetric to the skin 202. Gives a force based on vibration. Thereby, a pseudo translational force sensation can be presented in any direction along the D-1 axis.

  Desirably, the response waveform (time-series waveform pattern) of the force applied to the skin 202 by the contact portion 1003-1 has a direction of the predetermined direction DIR1 and an absolute value in the time interval τ1 (first time interval). The force is equal to or greater than the threshold TH1 (first threshold), and in the time interval τ2 (second time interval different from the first time interval), the direction is the opposite direction DIR2 of the predetermined direction and the absolute value is the threshold TH2. It represents a force within (TH2 <TH1). However, τ1 <τ2, and the time interval τ1 and the time interval τ2 are periodically repeated. The target response waveform pattern that realizes such a response waveform represents a force whose direction is a predetermined direction and whose absolute value is greater than or equal to a first threshold in the first time interval T1, and the first time interval T1 The different second time period T2 is a periodic waveform pattern representing a force whose direction is opposite to the predetermined direction and whose absolute value is within the second threshold. However, the first time interval T1 is shorter than the second time interval T2. As a result, a pseudo force sense can be perceived more clearly. More desirably, the waveform pattern of the force is a rectangular pattern or a pattern that approximates a rectangular pattern.

[Modification 1 of the seventh embodiment]
In the seventh embodiment, the D-1 axis is an axis along the short direction of the base portion 1001, but as illustrated in FIGS. 17A and 17B, the axis along the longitudinal direction of the base portion 1001 is the D-1 axis. It is good.

[Modification 2 of the seventh embodiment]
In the seventh embodiment, the support portion 1026-1 of the vibrator 1020-1 is attached to the bottom surface 1001e-1 of the recess 1001d-1 of the base portion 1001, and the contact portion 1003- is connected to the moving member 1025-1 of the vibrator 1020-1. 1 was attached. However, as illustrated in FIG. 17C, the moving member 1025-1 of the vibrator 1020-1 is attached to the bottom surface 1001e-1 of the recess 1001d-1 of the base 1001 via the connecting portion, and the vibrator 1020-1 is supported. The contact part 1003-1 may be attached to the part 1026-1. In this case, the “base mechanism” is a mechanism including the base 1001 and the motion member 1025-1. For example, the base 1001, the motion member 1025-1, and their connecting portions 1029a-1, 1029b-1, 1029c-1, 1029d-1. Further, the “contact mechanism” in this case is a mechanism including the contact portion 1003-1 and the support portion 1026-1, for example, a mechanism including the contact portion 1003-1 and the support portion 1026-1. The “contact mechanism” may further include at least one of the springs 1022-i, 1023-i, and the coil 1024-i. The mass m ₁ of the “contact mechanism” is smaller than the mass m ₂ of the “base mechanism”. This effectively presents a pseudo force sense. Preferably, the mass m ₁ of the “contact mechanism” is greater than zero and not more than one third of the mass of the mass m ₂ of the “base mechanism”.

[Modification 3 of the seventh embodiment]
The simulated force sense generation device 1000 may not include the base 1001. For example, the mechanism including the vibrator 1020-1 and the contact part 1003-1 and the base part 1001 circulate separately, and when used, the vibrator 1020-1 is connected to the base part 1001 as in Modification 1 or 2 of the seventh embodiment. And the contact part 1003-1 may be attached.

When attached as in the seventh embodiment (FIG. 15A), the “base mechanism” is a mechanism including the support portion 1026-1, for example, a mechanism including the support portion 1026-1. The “base mechanism” may further include at least one of springs 1022-i, 1023-i, and coils 1024-i. The “contact mechanism” is a mechanism including the contact portion 1003-1 and the motion member 1025-1. For example, the contact portion 1003-1, the motion member 1025-1, and the connection portions 1029a-i, 1029b-i, and 1029c- i, 1029d-i. The mass m ₁ of the “contact mechanism” is smaller than the total m ₂ of the mass of the “base mechanism” and the mass of the base 1001 that is a mechanism attached to the “base mechanism”. In dynamic characteristic model Md in this case, "contact mechanism" corresponds to the mass point M _1, a base 1001 as "base mechanism" corresponds to the mass point M _2.

When attached like the second modification of the seventh embodiment (FIG. 17C), the “base mechanism” is a mechanism including the motion member 1025-1, for example, the motion member 1025-1 and their connecting portions 1029a−. 1, 1029b-1, 1029c-1, and 1029d-1. The “contact mechanism” in this case is a mechanism including the contact portion 1003-1 and the support portion 1026-1, and is a mechanism including the contact portion 1003-1 and the support portion 1026-1, for example. The “contact mechanism” may further include at least one of the springs 1022-i, 1023-i, and the coil 1024-i. The mass m ₁ of the “contact mechanism” is smaller than the total m ₂ of the mass of the “base mechanism” and the mass of the base 1001 that is a mechanism attached to the “base mechanism”. In dynamic characteristic model Md in this case, "contact mechanism" corresponds to the mass point M _1, a base 1001 as "base mechanism" corresponds to the mass point M _2.

In these cases also, preferably 0 <m ₁ / m ₂ ≦ 1/3.

[Eighth Embodiment]
The eighth embodiment is a modification of the seventh embodiment, and the simulated force sense generator has a plurality of contact mechanisms.

  As illustrated in FIG. 18A to FIG. 18D, the pseudo force generation device 1100 according to the present embodiment is a motion member that performs asymmetric vibration (periodic asymmetric motion) with respect to the base 1101, the support 1026-i, and the support 1026-i. The vibrator 1020-i including 1025-i and the contact portion 1003-i are provided (where i = 1, 2). The difference from the seventh embodiment is that not only the contact portion 1003-1 is disposed on the upper surface 1101b side of the base portion 1101, but also the contact portion 1003-2 is disposed on the lower surface 1101c side of the base portion 1101. It is. Each contact portion 1003-i is attached to the base portion 1101 via the vibrator 1020-i, and the contact portion 1003-i asymmetrically vibrates with respect to the base portion 1101 by driving the vibrator 1020-i. As in the seventh embodiment, the support portion 1026-i may be attached to the base portion 1101, the vibrator 1020-i may be attached to the contact portion 1003-i, or Modification 2 of the seventh embodiment. As described above, the vibrator 1020-i may be attached to the base part 1101, and the support part 1026-i may be attached to the contact part 1003-i. However, both the D-1 axis and the D-2 axis in this embodiment are axes along the short direction of the base 1101 and are substantially parallel to each other. The contact portion 1003-1 is disposed substantially parallel to the upper surface 1101b, and the contact portion 1003-2 is disposed substantially parallel to the lower surface 1101c. In this example, the contact portion 1003-1 and the contact portion 1003-2 are disposed substantially symmetrically with respect to the intermediate surface between the upper surface 1101b and the lower surface 1101c of the base portion 1101. That is, the vibrators 1020-1 and 1020-2 are arranged between the contact part 1003-1 and the contact part 1003-2.

  When using the pseudo force sense generating apparatus 1100, the base 1101 and the contact parts 1003-1 and 1003-2 come into contact with the user's skin. As illustrated in FIG. 18C, when the contact parts 1003-1 and 1003-2 are asymmetrically vibrated so as to present pseudo force sensations in the same directions E-1 and E-2, the user moves in that direction. Perception of translational force sense. On the other hand, as illustrated in FIG. 18C, when the contact portions 1003-1 and 1003-2 are asymmetrically vibrated so as to present pseudo force senses in opposite directions E-1 and E-2, the user A pseudo force sense that rotates in the E-11 direction is perceived.

[Modification 1 of Eighth Embodiment]
Like the simulated force sense generator 1200 illustrated in FIGS. 19A to 19C, the D-1 axis and the D-2 axis may be substantially orthogonal to each other. An example of “substantially orthogonal” is orthogonal. The pseudo force sense generator 1200 includes a base 1201, a support 1026-i, a vibrator 1020-i including a moving member 1025-i that performs asymmetric vibration (periodic asymmetric motion) with respect to the support 1026-i, and a contact. Part 1003-i (where i = 1, 2). A contact portion 1003-1 is disposed on the upper surface 1201b side of the base portion 1201, and a contact portion 1003-2 is disposed on the lower surface 1201c side of the base portion 1201. Each contact portion 1003-i is attached to the base portion 1201 via the vibrator 1020-i, and the contact portion 1003-i asymmetrically vibrates with respect to the base portion 1201 by driving the vibrator 1020-i. The D-1 axis is an axis along the short direction of the base portion 1201, and the D-2 axis is an axis along the longitudinal direction of the base portion 1201, and is substantially orthogonal to each other. The contact portion 1003-1 is disposed substantially parallel to the upper surface 1201b, and the contact portion 1003-2 is disposed substantially parallel to the lower surface 1201c. In this example, the contact part 1003-1 and the contact part 1003-2 are arranged at positions shifted from each other. That is, the contact portion 1003-2 does not exist in a region that is plane-symmetric with the contact portion 1003-1 with respect to an intermediate surface that is substantially parallel to the upper surface 1201 b and the lower surface 1201 c. When the simulated force sense generator 1200 is used, the base 1201 and the contact parts 1003-1 and 1003-2 come into contact with the user's skin. By driving the vibrators 1020-1 and 1020-2 of the simulated force sense generator 1200, the user perceives a pseudo force sense based on the asymmetric vibration of the contact parts 1003-1 and 1003-2.

[Modification 2 of Eighth Embodiment]
Like the pseudo force generation device 1300 illustrated in FIGS. 20A to 20C, the contact portion 1003-1 is plane-symmetric with respect to an intermediate surface substantially parallel to the upper surface 1201b and the lower surface 1201c. The contact portion 1003-2 may exist in the region. When such a pseudo force generation device 1300 is used, the base 1201 and the contact parts 1003-1 and 1003-2 come into contact with the user's skin. By driving the vibrator 1020-i of the simulated force sense generator 1200 and driving the vibrators 1020-1 and 1020-2 of the simulated force sense generator 1300, the user can make contact portions 1003-1 and 1003. -2 senses a pseudo force sense based on -2 asymmetric vibration.

[Modification 3 of Eighth Embodiment]
The simulated force sense generator 1400 illustrated in FIGS. 21A to 21C includes a base portion 1401, a vibrator 1020-i, and a contact portion 1003-i (where i = 1, 2). One side of the support portion 1026-2 or the moving member 1025-2 of the vibrator 1020-2 is attached to the inside of the base portion 1401, and the other side is attached to the contact portion 1003-2. One side of the support part 1026-1 of the vibrator 1020-1 or the moving member 1025-1 is attached to the contact part 1003-2, and the other side is attached to the contact part 1003-1. . The vibrator 1020-1 is attached such that the moving member 1025-1 can asymmetrically vibrate along the D-1 axis along the short side direction of the base 1401. The member 1025-2 is attached so that it can asymmetrically vibrate along the D-2 axis along the longitudinal direction of the base 1401. That is, the vibrators 1020-1 and 1020-2 are attached such that the D-1 axis and the D-2 axis are substantially orthogonal to each other. For example, as illustrated in FIGS. 22A and 22B, the support portion 1026-2 of the vibrator 1020-2 is fixed inside the base portion 1401, and the motion member 1025-2 of the vibrator 1020-2 is connected to the connecting portion 1029a. It is attached to the contact portion 1003-2 through −2, 1029 b-2, 1029 c-2, and 1029 d-2. The contact portion 1003-2 is attached with a support portion 1026-1 of the vibrator 1020-1, and the moving member 1025-1 of the vibrator 1020-1 is connected to the connecting portions 1029 a-1, 1029 b-1, and 1029 c-1. , 1029d-1 to the contact portion 1003-1. The contact portions 1003-1 and 1003-2 are substantially parallel to the upper surface 1401b and the lower surface 1401c of the base portion 1401, and the contact portion 1003-1 is disposed on substantially the same plane as the upper surface 1401b.

  When the simulated force sense generator 1400 is used, the base portion 1401 and the contact portion 1003-1 come into contact with the user's skin. When only the vibrator 1020-1 is driven, a pseudo force sense can be presented in any direction along the D-1 axis, and when only the vibrator 1020-2 is driven, D -A pseudo force sense can be presented in any direction along the -2 axis. When both the vibrator 1020-1 and the vibrator 1020-2 are driven, a pseudo force sense can be presented in an arbitrary direction on a plane composed of the D-1 axis and the D-2 axis.

[Ninth Embodiment]
As illustrated in FIG. 23A and FIG. 23B, the simulated force generation device 1500 of this embodiment includes a vibrator 1020-1, a plate-like contact portion 1503, and a belt-like (band-like) contact portion 1504. The contact portion 1503 is made of synthetic resin or the like, and the contact portion 1504 is made of synthetic resin or leather. The moving member 1025-1 of the vibrator 1020-1 is attached to the plate surface 1503c of the contact portion 1503 via the connecting portions 1029a-1, 1029b-1, 1029c-1, and 1029d-1. Further, the contact portion 1504 is attached to the support portion 1026-1 of the vibrator 1020-1. In the example of FIGS. 23A and 23B, a belt-like contact portion 1504 is attached to the support portion 1026-1 in a direction substantially orthogonal to the D-1 axis. That is, the contact portion 1504 is attached to the support portion 1026-1 in a direction substantially orthogonal to the D-1 axis. As illustrated in FIG. 23B, the simulated force sense generator 1500 is mounted such that the contact portions 1503 and 1504 are in contact with the skin of the user's arm 1510, for example.

In the case of this configuration, the “base mechanism” is a mechanism including the support portion 1026-1 and the contact portion 1504. For example, the “base mechanism” is a mechanism including the support portion 1026-1 and the contact portion 1504. The “base mechanism” may further include at least one of springs 1022-i, 1023-i, and coils 1024-i. The “contact mechanism” is a mechanism including the contact portion 1503 and the moving member 1025-1. For example, the contact portion 1503-1, the moving member 1025-1, and the connecting portions 1029a-i, 1029b-i, 1029c-i, 1029d-i. The mass m ₁ of the “contact mechanism” is smaller than the mass m ₂ of the “base mechanism”. Preferably, 0 <m ₁ / m ₂ ≦ 1/3. In the dynamic characteristic model Md in this case, the “contact mechanism” corresponds to the mass point M ₁ and the “base mechanism” corresponds to the mass point M ₂ . Thereby, a pseudo force sense can be presented in any direction along the direction in which the arm 1510 extends.

[Modification 1 of Ninth Embodiment]
24, a belt-like member 1604 made of synthetic resin, leather, or the like is attached to the support 1026-1, and the belt-like member 1604 is further along the D-1 axis. The strip-shaped contact portion 1504 may be attached in an annular shape in the opposite direction. That is, the contact portion 1504 is attached to the support portion 1026-1 in a direction along the D-1 axis. In the case of this configuration, the “base mechanism” is a mechanism including the support portion 1026-1, the strip-shaped member 1604, and the contact portion 1504. For example, a mechanism including the support portion 1026-1, the strip-shaped member 1604, and the contact portion 1504. It is. The “base mechanism” may further include at least one of springs 1022-i, 1023-i, and coils 1024-i. The “contact mechanism” is a mechanism including the contact portion 1503 and the moving member 1025-1. For example, the contact portion 1503-1, the moving member 1025-1, and the connecting portions 1029a-i, 1029b-i, 1029c-i, 1029d-i. The mass m ₁ of the “contact mechanism” is smaller than the mass m ₂ of the “base mechanism”. In this configuration, a pseudo force sense can be presented in any direction across the arm 1510.

[Tenth embodiment]
In the case where both the contact mechanism and the base mechanism are in contact with the skin (for example, the seventh to ninth embodiments and their modifications), even if the contact mechanism performs asymmetric vibration according to the target control signal waveform and the bias component. Good. However, the bias component is a component that controls the vibration center of the asymmetric vibration. That is, the position of the vibration center of the asymmetric vibration of the contact mechanism may be controllable. For example, the vibration center may be moved in a direction in which a pseudo force sense due to asymmetric vibration is presented by the bias component. Thereby, a pseudo force sense can be perceived more clearly.

For example, as illustrated in FIG. 25A, along the Di axis of the contact portion 1003-i when the vibrator 1020-i is driven according to only the target control signal waveform (when the bias component is zero). Let Φ be the vibration center of the asymmetric vibration. Here, in the case of presenting a pseudo force to the G1 direction as illustrated in FIG. 25B, the vibration center of the asymmetric oscillation of the contact portion 1003-i is moved in the G1 direction than the vibration center [Phi [Phi ₁ and The bias component is adjusted so that For example, when a force in the G1 direction is applied to the contact portion 1003-i when the target control signal waveform is positive, a control signal corresponding to the waveform obtained by adding the target control signal waveform and the positive bias component is an electric circuit. 13 and a corresponding drive signal is applied to the vibrator 1020-i. Conversely, when a pseudo force sense is presented in the G2 direction as illustrated in FIG. 25C, the vibration center of the asymmetric vibration of the contact portion 1003-i is Φ _{2 that} is moved in the G2 direction from the vibration center Φ. The bias component is adjusted as follows. For example, if a force in the G2 direction is applied to the contact portion 1003-i when the target control signal waveform is negative, a control signal corresponding to a waveform obtained by subtracting a positive bias component from the target control signal waveform is supplied to the electric circuit 13. And a corresponding drive signal is applied to the vibrator 1020-i. Note that the magnitude of the bias component for moving the vibration center of the asymmetric vibration of the contact portion 1003-i in the G1 direction and the magnitude of the bias component for moving the vibration center in the G2 direction are the same. Or may not be the same. The magnitude of the bias component for moving the vibration center in the G1 direction or the bias component for moving the vibration center in the G2 direction according to the past history of the pseudo force sense presentation direction. You may change the magnitude | size of. Note that in the examples in FIGS. 25B and 25C, the G1 direction and the G2 direction are in a direction that antagonizes, so the bias at a certain time is either the G1 direction or the G2 direction.

[simulation result]
The comparison simulation result is shown.
<Comparison between the pseudo force sense generator 12 and the pseudo force sense generator 1000>
FIG. 26A to FIG. 26C and FIGS. 27A to 27C show comparison simulation results when control is performed with a time-asymmetric rectangular wave control voltage. However, this time-asymmetric rectangular wave has a positive period T1 and a negative period T2 of [T1, T2] = [6, 12] [ms]. FIGS. 26A to 26C show simulation results when the pseudo force sense generator 12 (FIG. 3A) is built in a main body of a smartphone terminal device or the like to which a contact portion such as a smartphone case is attached. Represents a simulation result in the pseudo force sense generator 1000 (FIG. 15A). FIG. 26A and FIG. 27A show control voltages of the pseudo force sense generator 12 and the pseudo force sense generator 1000. The vertical axis represents the voltage value [V] of the control voltage, and the horizontal axis represents time [sec]. 26B and 27B show a contact portion attached to a main body in which the pseudo force sense generator 12 is built and a contact portion 1003 of the pseudo force sense generator 1000 when driven by the control voltage of FIGS. 26A and 27A. -1 represents the force applied to the skin. The vertical axis represents force [N] applied to the skin, and the horizontal axis represents time [sec]. 26C and 27C show a contact portion attached to the main body in which the pseudo force sense generator 12 is built and a contact portion 1003 of the pseudo force sense generator 1000 when driven by the control voltage of FIGS. 26A and 27A. -1 position (vibration waveform). The vertical axis represents the position [m] of the contact portion, and the horizontal axis represents time [sec]. However, in FIGS. 26A to 26C, the mass of a system including a main body (for example, a smartphone terminal device) with a mass of 145 g, a contact portion (for example, a smartphone case) with a mass of 10 g, and a support portion (for example, an actuator case) with a mass of 10 g. m ₁ = 155 g, and the mass of a system composed of a moving member (for example, an actuator mover) having a mass of 5 g was m ₂ = 5 g. On the other hand, in FIGS. 27A to 27C, the mass of the system including the contact portion 1003-1 having a mass of 5 g and the moving member 1025-i having a mass of 5 g (for example, an actuator mover) is m ₁ = 10 g, and the support portion having a mass of 10 g is used. The mass of a system consisting of 1026-i (for example, an actuator case) and a base 1001 (for example, a smartphone terminal device) having a mass of 135 g was set to m ₂ = 145 g. As can be seen from these figures, the pseudo force generation device 1000 can increase the mass m ₂ with respect to the mass m ₁ as compared to the configuration in which the pseudo force generation device 12 is built in the main body to which the contact portion is attached. Both the asymmetry of the vibration of the contact portion and the asymmetry of the force applied to the skin can be increased. As a result, the configuration of this embodiment can present a pseudo force sense more clearly than in the past.

<Comparison 1 between control voltage of time asymmetric rectangular wave and control voltage of target control signal waveform>
28A to 28C show simulation results when the pseudo force sense generator 1000 is controlled with a time asymmetric rectangular wave control voltage, and FIGS. 29A to 29C show the pseudo force sense generator 1000 with the control voltage of the target control signal waveform. The comparison simulation result when controlled by is shown. However, the time asymmetric rectangular wave and the target control signal waveform are both those in which the positive period T1 and the negative period T2 are [T1, T2] = [3, 16] [ms]. The configuration of the simulated force generation apparatus 1000 is the same as the configuration of FIGS. 27A to 27C. FIG. 28A and FIG. 29A show the control voltage of the simulated force sense generator 1000. The vertical axis represents the voltage value [V] of the control voltage, and the horizontal axis represents time [sec]. FIGS. 28B and 29B show the force applied to the skin from the contact portion 1003-1 of the simulated force sense generator 1000 when driven by the control voltage of FIGS. 28A and 29A. The vertical axis represents force [N] applied to the skin, and the horizontal axis represents time [sec]. However, the broken line in FIG. 29B represents the target response waveform pattern, and the solid line represents the response waveform. 28C and 29C show the position of the contact portion 1003-1 of the simulated force sense generator 1000 when driven with the control voltage of FIGS. 28A and 29A. The vertical axis represents the position [m] of the contact portion, and the horizontal axis represents time [sec]. As can be seen from these figures, when the control is performed with the control voltage of the target control signal waveform, both the force applied to the skin and the position of the contact portion 1003-1 become noise, compared with the case of control with the control voltage of the time asymmetric rectangular wave. There is little fine vibration of the waveform, while the difference between the maximum value in the positive direction and the maximum value in the negative direction is large, and the asymmetry is large. From this, it can be seen that a clearer sense of force can be presented by controlling with the control voltage of the target control signal waveform.

<Comparison 2 between control voltage of time asymmetric rectangular wave and control voltage of target control signal waveform>
30A to 30C show simulation results when the pseudo force sense generator 1000 is controlled with a time-asymmetric rectangular wave control voltage, and FIGS. 31A to 31C show the pseudo force sense generator 1000 with the control voltage of the target control signal waveform. The comparison simulation result when controlled by is shown. However, the time asymmetric rectangular wave and the target control signal waveform are those in which the positive period T1 and the negative period T2 are [T1, T2] = [6, 12] [ms]. Other conditions are the same as those in <Comparison 1 between control voltage of time asymmetric rectangular wave and control voltage of target control signal waveform>. Also from these figures, it can be seen that a clearer sense of force can be presented by controlling with the control voltage of the target control signal waveform.

<Comparison 3 between control voltage of time asymmetric rectangular wave and control voltage of target control signal waveform>
FIG. 32A shows an example of the asymmetry of the force applied to the skin from the contact portion 1003-1 when the pseudo force generation device 1000 is controlled by the control voltage of the time asymmetric rectangular wave of [T1, T2] [ms]. It is the figure which carried out the stem plot. FIG. 32B shows an example of asymmetry of the force applied to the skin from the contact portion 1003-1 when the pseudo force generation device 1000 is controlled with the control voltage of the target control signal waveform of [T1, T2] [ms]. It is the figure which carried out the stem plot. The two axes at the bottom of each figure represent the periods T1 and T2, respectively, and the vertical axis represents the difference between the positive and negative large forces that are the asymmetry of the force applied to the skin (more precisely, the maximum value in the positive direction and the negative value in the negative direction). Sum of local minimum values, ie, the absolute value of the local maximum value in the positive direction minus the absolute value of the local minimum value in the negative direction.) [N]. Other conditions are the same as those in <Comparison 1 between control voltage of time asymmetric rectangular wave and control voltage of target control signal waveform>. From these figures, the asymmetry of the force applied to the skin from the contact portion 1003-1 is increased by controlling with the control voltage of the target control signal waveform as compared with the case of controlling with the control voltage of the time asymmetric rectangular wave. I understand. Furthermore, it can be seen that by controlling with the control voltage of the target control signal waveform, there is a robust tendency against changes in [T1, T2].

[Other variations]
The present invention is not limited to the embodiment described above. For example, the configuration of the pseudo force sense generator is not limited to the above-described configuration, and may be another pseudo force sense generator that generates a pseudo force sense by periodic asymmetric motion. The configuration of the gripping part is not limited to the above. That is, the gripping part does not have to be the case of the pseudo force sense generator itself, but may be a grip part mechanically fixed to the pseudo force sense generator. Further, the simulated force sense generator is not limited to that of the above-described embodiment or its modification. For example, a period during which an acceleration measurement result by an acceleration sensor installed in the simulated force sense generator is equal to or less than a predetermined threshold is a predetermined value. If it is longer than the length, it may be determined that the pseudo force generation device is not gripped, and an additional function of operating in a different mode may be provided. When such a function is added, for example, the pseudo force sense presentation may be stopped when the user is not gripped, and further, the fourth modification of the first and second embodiments may not be performed. By stopping the process when not gripped, it is possible to reduce the risk of dropping or damage due to the operation of the apparatus without gripping, or battery consumption.

The equations of motion of equations (14)-(17) can be generalized even when j_max = 2. The mathematical expressions of the mechanical property model Md and the skin mechanical property model Ms may be as follows.

  An example of “periodic asymmetric motion” is asymmetric vibration. Further, the “skin” in each embodiment and its modifications may be replaced with “mucosa”.

  In each embodiment, as an example of the “trial signal waveform” and the “target control signal waveform”, the time waveform of the voltage and the time waveform of the current are exemplified, but the “trial signal waveform” and the “target control signal waveform” are these. It is not limited to. For example, the waveform may be a waveform represented by a discontinuous function defined by several parameters.

  Examples of waveforms represented by discontinuous functions are shown in FIGS. 33A and 33B. FIG. 33A is an example in which a waveform is defined by a total of seven parameters. This waveform shows TA-TD indicating the length of the time interval, AM1 and AM2 indicating the voltage or current value (voltage value or current value), and the slope of the linear function indicating the voltage or current value in the time interval TD. A total of seven dAMs (where the linear function is a function of a time variable, and the function value at the end of the time interval immediately before the time interval TD is the voltage or current value at the end) It is expressed by the parameter. FIG. 33B is an example when the length of the time interval of TC is zero.

For example, a “trial signal waveform” and a “target control signal waveform” having the following characteristics in one cycle (a time period from the start time of TA to the end time of TD, represented by T in FIG. 33A), This is considered to be an effective waveform for presenting pseudo force sense.
<< Characteristic 1 >> This is a time interval in which a value after change is maintained after an "acceleration period" in which the value of voltage or current or the direction of current is suddenly changed to accelerate the moving member 1025-i in a certain direction. There is a “duration” (eg, TB in FIG. 33A).
<< Characteristic 2 >> After the "duration" (more preferably, immediately after), there is a "deceleration period" (for example, TC in FIG. 33A), which is a time period during which the motion member 1025-i is decelerated rapidly. For example, when the sign (positive / negative) of the voltage value is reversed between the “duration period” and the “deceleration period” and the voltage value in the “deceleration period” is changed to be as large as possible, the moving member 1025-i is suddenly changed. The effect of slowing down can be expected. In the case of current, if the direction of current flow is reversed in the “duration period” and the “deceleration period” and the amount of current in the “deceleration period” is changed to be as large as possible, the motion member 1025 is abruptly changed. The effect of decelerating -i can be expected.
<< Characteristic 3 >> After the "deceleration period" (more preferably, immediately after), the acceleration applied to the exercise member 1025-i to return the position of the exercise member 1025-i to the position before acceleration in the "acceleration period" There is a “return period” (TD in FIG. 33A) in which (or force) is gradually changed in the direction opposite to the acceleration direction in the “acceleration period”. For example, in the “recovery period”, the voltage value may be changed according to a linear function. However, even a waveform that does not have “feature 2” as shown in FIG. 33B may be a waveform that is effective for presenting a pseudo force sense.

  Further, the “trial signal waveform” and the “target control signal waveform” may be discrete data that takes a predetermined value at each discrete time (for example, a predetermined sampling time).

  As illustrated in FIG. 34, a control signal estimation device 11 ′ (or 21 ′) including each unit included in the control signal estimation device 11 (or 21) described above and a target control signal simplification unit 516 is used. Also good. The target control signal simplification unit 516 receives a waveform (referred to as “inverse dynamics waveform”) obtained by applying the target response waveform pattern to the inverse dynamics model Mc in the model application unit 113 (or 213), and performs the inverse process. A waveform obtained by simplifying the dynamics waveform is output as a target control signal waveform.

  For example, the target control signal simplification unit 516 receives the inverse dynamics waveform from the model application unit 113 (or 213) and converts it into a parameter group (for example, seven parameters TA to TD, AM1, AM2, and dAM in FIG. 33A). A target control signal waveform expressed by a waveform pattern defined by (for example, FIG. 33A) is obtained, and information indicating the waveform pattern and / or a parameter group corresponding thereto is output. The electric circuit 13 operates as described in the above-described embodiments in accordance with the information representing the waveform pattern and / or the target control signal waveform specified by the corresponding parameter group. Here, the data amount of the parameter group and the data amount of the waveform pattern defined by the parameter group are smaller than the data amount of the inverse dynamics waveform. Thereby, the storage capacity when storing information for specifying the target control signal waveform and the communication capacity when transmitting this information can be reduced.

  For example, the target control signal simplification unit 516 obtains the value of the parameter group of the waveform pattern that minimizes the error between the inverse dynamics waveform and the waveform pattern defined by the predetermined parameter group by the least square method, Such a waveform pattern may be used as the target control signal waveform. For example, the target control signal simplification unit 516 has the smallest error (for example, the sum of squares of errors in a predetermined time interval) between the inverse dynamics waveform at each time and the waveform pattern defined by the above parameter group. Such parameter group values may be obtained by the method of least squares.

  There may be a plurality of groups of waveform pattern types defined by a predetermined parameter group. For example, a waveform pattern group defined by a parameter group including seven parameters TA to TD, AM1, AM2, and dAM in FIG. 33A, and a parameter group of a different type (for example, a parameter not having the above-described feature 2). There may be a group of waveform patterns defined in (Group). In this case, the target control signal simplification unit 516 may select the waveform pattern closest to the inverse dynamics waveform from among the waveform patterns defined by the parameters of each group, and use the selected waveform pattern as the target control signal waveform. . For example, the target control signal simplification unit 516 selects, as the target control signal waveform, a waveform pattern that minimizes an error from the inverse dynamics waveform among the waveform patterns defined by the parameter groups of each group, and the waveform pattern And a value of a parameter group defining the information may be output as “target control signal waveform” information.

  Note that, as illustrated in FIG. 34, the control signal estimation device 11 ′ (or 21 ′) may further include a sampling unit 517. The sampling unit 517 receives the waveform pattern selected by the target control signal simplification unit 516 and samples the waveform pattern value at a predetermined sampling time, thereby obtaining waveform information (for example, a waveform at each sampling time). List of values) may be created, and the created waveform information may be output as a “target control signal waveform”.

  As illustrated in FIG. 35, a control signal estimation device 51 ′ (or 61 ′) having each unit included in the control signal estimation device 51 (or 61) and a target control signal simplification unit 516 ′ is used. May be. The target control signal simplification unit 516 ′ receives the latest trial signal waveform TR (t) (adjusted trial signal waveform obtained by adjusting the trial signal waveform) that satisfies the termination condition in the trial signal update unit 514, as input, A waveform obtained by simplifying the trial signal waveform TR (t) is output as a target control signal waveform.

  For example, the target control signal simplification unit 516 ′ receives the latest trial signal waveform TR (t) that satisfies the termination condition from the trial signal update unit 514, and uses this as a parameter group (for example, TA to TD, AM1 in FIG. 33A). , AM2, and dAM)), a target control signal waveform expressed by a waveform pattern (for example, FIG. 33A) is obtained, and information indicating the waveform pattern and / or a parameter group corresponding thereto are output. The electric circuit 13 operates as described in the above-described embodiments in accordance with the information representing the waveform pattern and / or the target control signal waveform specified by the corresponding parameter group. Here, the data amount of the parameter group and the data amount of the waveform pattern defined by the parameter group are smaller than the data amount of the latest trial signal waveform TR (t) that satisfies the termination condition. Thereby, the storage capacity when storing information for specifying the target control signal waveform and the communication capacity when transmitting this information can be reduced.

  For example, the target control signal simplification unit 516 ′ uses a waveform pattern that minimizes the error between the latest trial signal waveform TR (t) that satisfies the termination condition and the waveform pattern defined by a predetermined parameter group. The parameter group values may be obtained by the least square method, and such a waveform pattern may be used as the target control signal waveform. For example, the target control signal simplification unit 516 ′ uses an error (for example, a predetermined value) between the latest trial signal waveform TR (t) that satisfies the termination condition at each time t and the waveform pattern defined by the above parameter group. The value of the parameter group that minimizes the sum of squares of errors in the time interval of (1) may be obtained by the method of least squares.

  There may be a plurality of groups of waveform pattern types defined by a predetermined parameter group. For example, a waveform pattern group defined by a parameter group including seven parameters TA to TD, AM1, AM2, and dAM in FIG. 33A, and a parameter group of a different type (for example, a parameter not having the above-described feature 2). There may be a group of waveform patterns defined in (Group). In this case, the target control signal simplification unit 516 ′ selects the waveform pattern defined by the parameter group of each group that is closest to the latest trial signal waveform TR (t) that satisfies the termination condition, The selected waveform pattern may be the target control signal waveform. For example, the target control signal simplification unit 516 ′ has a waveform pattern that minimizes an error from the latest trial signal waveform TR (t) that satisfies the termination condition among the waveform patterns defined by the parameters of each group. May be selected as the target control signal waveform, and the information indicating the waveform pattern and the value of the parameter group defining it may be output as the “target control signal waveform” information.

  The target control signal simplification unit 516 ′ does not obtain the target control signal waveform by simplifying the latest trial signal waveform TR (t) that satisfies the termination condition by the trial signal update unit 514, but the control signal estimation device described above. The 51 (or 61) trial signal updating units 514 may directly output the trial signal waveform TR (t) including the waveform pattern defined by the parameter group as the target control signal waveform. That is, the processing of the above-described steps S52 to S56 is performed using the waveform pattern defined by the parameter group as the trial signal waveform TR (t), and the latest trial signal waveform TR (t) satisfying the end condition is the target control signal waveform. May be output. Also in this case, there may be a plurality of groups of waveform pattern types defined by a predetermined parameter group.

  Note that, as illustrated in FIG. 35, the control signal estimation device 51 ′ (or 61 ′) may further include a sampling unit 517. The sampling unit 517 receives the waveform pattern selected by the target control signal simplification unit 516 ′, and samples the waveform pattern value at a predetermined sampling time to obtain waveform information (for example, at each sampling time). A list of waveform values) may be created, and the created waveform information may be output as a “target control signal waveform”.

  When the above-described control signal estimation apparatus and waveform estimation apparatus are realized by a computer, the processing contents of functions that each apparatus should have are described by a program. By executing this program on a computer, the above processing functions are realized on the computer. The program describing the processing contents can be recorded on a computer-readable recording medium. An example of a computer-readable recording medium is a non-transitory recording medium. Examples of such a recording medium are a magnetic recording device, an optical disk, a magneto-optical recording medium, a semiconductor memory, and the like.

  This program is distributed, for example, by selling, transferring, or lending a portable recording medium such as a DVD or CD-ROM in which the program is recorded. Furthermore, the program may be distributed by storing the program in a storage device of the server computer and transferring the program from the server computer to another computer via a network.

  A computer that executes such a program first stores, for example, a program recorded on a portable recording medium or a program transferred from a server computer in its own storage device. When executing the process, the computer reads a program stored in its own storage device, and executes a process according to the read program. As another execution form of the program, the computer may read the program directly from the portable recording medium and execute processing according to the program, and each time the program is transferred from the server computer to the computer. The processing according to the received program may be executed sequentially. The above-described processing may be executed by a so-called ASP (Application Service Provider) type service that realizes a processing function only by an execution instruction and result acquisition without transferring a program from the server computer to the computer. Good.

  In the above embodiment, the processing functions of the control signal estimation device and the waveform estimation device are realized by executing a predetermined program on a computer. However, at least a part of these processing functions may be realized by hardware.

11, 21, 11 ', 21', 51, 61, 51 ', 61' Control signal estimation device 12, 12 ', 12 "Pseudo force sense generator 31, 41 Waveform estimation device

Claims (20)

  A dynamic characteristic model of a pseudo force sensation generator having a moving member that performs a periodic asymmetric motion according to a control signal waveform, and a gripping portion to which a force generated by the asymmetrical motion is applied, and a contact with the gripping portion Between the control signal waveform and a waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the skin, obtained based on a plurality of models including a skin dynamic characteristic model An estimated waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the skin, obtained by applying a trial signal waveform as the control signal waveform to a forward dynamics model of a control target representing And an error from a target response waveform pattern corresponding to at least one of the position, velocity, acceleration, and force applied to the skin Based adjust the trial signal waveform to obtain the target control signal waveform for controlling the asymmetric motion, control signal estimator. The control signal estimation device according to claim 1,
The forward dynamics model of the control object includes a dynamic characteristic model of the simulated force sense generator, a dynamic characteristic model of the skin, and an output of an electric circuit that supplies a drive signal to the control signal waveform and the simulated force sense generator. A control signal estimation device obtained on the basis of an electric circuit characteristic model that represents the relationship. The control signal estimation device according to claim 1 or 2,
The pseudo force sense generating device includes:
The gripping part;
A support portion having a fixed relative position to the gripping portion;
An elastic body supported by the support part;
The motion member supported by the elastic body and performing a periodic asymmetric motion with respect to the support portion;
A coil for applying a force corresponding to the control signal waveform to the moving member,
The dynamic system model of the pseudo force sense generator is:
Modeling the relationship between the force applied to the moving member by the coil, the position and speed corresponding to the gripping part, and the position and speed corresponding to the moving member,
The skin dynamic system model is:
A control signal estimation device that models a relationship between a position and a speed corresponding to the grip portion and a force applied to the skin. The control signal estimation device according to any one of claims 1 to 3,
The target response waveform pattern is:
The first time interval represents a force or acceleration whose direction is a predetermined direction and whose absolute value is greater than or equal to a first threshold,
In a second time interval different from the first time interval, the direction is opposite to the predetermined direction, and the absolute value is within the first threshold value or within a second threshold value that is smaller than the first threshold value. Represents a force or acceleration,
The control signal estimation device, wherein the first time interval is shorter than the second time interval. The control signal estimation device according to any one of claims 1 to 4,
A control signal estimation device that obtains the target control signal waveform obtained by expressing an adjusted trial signal waveform obtained by adjusting the trial signal waveform based on the error by a waveform pattern defined by a predetermined parameter group. A moving member that performs a periodic asymmetric motion according to a target control signal waveform obtained by the control signal estimation device according to claim 1,
A gripping portion to which a force generated by the asymmetric motion is applied;
A pseudo force sensation generating apparatus. The pseudo force sense generator according to claim 6,
The target control signal waveform is obtained based on a plurality of models including a mechanical characteristic model of the pseudo force sense generator and a mechanical characteristic model of the skin in contact with the grip part. , Obtained by applying a trial signal waveform as the control signal waveform to a forward dynamics model of a control target representing a relationship between the waveform corresponding to at least one of acceleration and force applied to the skin and the control signal waveform, An estimated waveform corresponding to at least one of the position, speed, acceleration, and force applied to the skin of the grip part, and at least one of the position, speed, acceleration, and force applied to the skin of the target grip part The pseudo force sense generator, which is the trial signal waveform adjusted based on an error from a corresponding target response waveform pattern. A first base mechanism supported directly or indirectly by a subject ;
Perform a periodic asymmetrical movement in response to a first control signal waveform which is a target control signal waveform relative to the first base mechanism, directly or indirectly contact with the first to the first skin of the subject was a first contact mechanism which gives the force based on the asymmetric movement in response to the control signal waveform, a pseudo force generator for have a,
The mass of the first contact mechanism is smaller than the mass of the first base mechanism, or the mass of the first contact mechanism is the mass of the first base mechanism and the mass of the mechanism attached to the first base mechanism. Less than the sum,
The target control signal waveform is
A second base mechanism that is directly or indirectly supported by a person, and a second second asymmetric motion corresponding to a second control signal waveform with respect to the second base mechanism, and a second second mechanism of the person A second contact mechanism that applies a force based on asymmetric motion to the skin , and a mechanical characteristic model of a second pseudo force sense generator, and a mechanical characteristic model of the second skin that contacts the second contact mechanism, A relationship between the second control signal waveform and a waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the second skin of the second contact mechanism obtained based on a plurality of models including At least one of the position, velocity, acceleration, and force applied to the second skin of the second contact mechanism, obtained by applying a trial signal waveform as the second control signal waveform to the forward dynamics model of the control target to be represented In The trial signal waveform is adjusted based on an error between a corresponding estimated waveform and a target response waveform pattern corresponding to at least one of the position, velocity, acceleration, and force applied to the skin of the second contact mechanism as a target A pseudo force generation device which is a waveform obtained in this manner.   The pseudo force sense generator according to claim 8,
  The second base mechanism is the first base mechanism;
  The second contact mechanism is the first contact mechanism;
  The pseudo force sense generating device, wherein the second asymmetric motion is the asymmetric motion. The simulated force sense generator according to claim 8 or 9 ,
The forward dynamics model of the control target includes a mechanical characteristic model of the second pseudo force sense generator, a mechanical characteristic model of the skin, and an electric power that supplies a drive signal to the control signal waveform and the second pseudo force sense generator. A pseudo force generation device obtained based on an electric circuit characteristic model representing a relationship with an output of a circuit. The pseudo force sense generator according to any one of claims 8 to 10 ,
The target response waveform pattern is:
The first time interval represents a force or acceleration whose direction is a predetermined direction and whose absolute value is greater than or equal to a first threshold,
In a second time interval different from the first time interval, the direction is opposite to the predetermined direction, and the absolute value is within the first threshold value or within a second threshold value that is smaller than the first threshold value. Represents a force or acceleration,
The pseudo force sense generating device, wherein the first time interval is shorter than the second time interval. The pseudo force sense generator according to any one of claims 8 to 11,
The simulated force sense generator, wherein the mass of the contact mechanism is greater than zero and less than or equal to one-third of the mass of the base mechanism. Be any of the pseudo force generator of claim 8 1 2,
The asymmetric motion is asymmetric vibration according to the target control signal waveform and a bias component,
The pseudo force sense generator, wherein the bias component is a component that controls a vibration center of the asymmetric vibration. A pseudo force generating device according to claim 1 3,
The pseudo force sense generation device, wherein the bias component is a component that moves the vibration center in a direction in which a pseudo force sense due to the asymmetric vibration is presented. Be any of the pseudo force generator of claim 8 1 4,
The target control signal waveform is
The pseudo force sense is a waveform obtained by adjusting the adjusted trial signal waveform obtained by adjusting the trial signal waveform based on the error, and further expressing the adjusted trial signal waveform by a waveform pattern defined by a predetermined parameter group. Generator. A dynamic characteristic model of a pseudo force sensation generator having a moving member that performs a periodic asymmetric motion according to a control signal waveform, and a gripping portion to which a force generated by the asymmetrical motion is applied, and a contact with the gripping portion Between the control signal waveform and a waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the skin, obtained based on a plurality of models including a skin dynamic characteristic model In the case where an input control signal waveform is applied as the control signal waveform to the forward dynamics model of the control target representing the position, the position, velocity, acceleration, and skin of the gripping part when the asymmetric motion is controlled by the input control signal waveform A waveform estimation device that estimates a response waveform corresponding to at least one of the forces applied to the.   A dynamic characteristic model of a pseudo force sensation generator to be modeled, which includes a moving member that performs a periodic asymmetric motion according to a control signal waveform, and a gripper to which a force generated by the asymmetric motion is applied, and A waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the skin, and the control signal, obtained based on a plurality of models including a mechanical characteristic model of the skin in contact with the grip Corresponds to at least one of the position, velocity, acceleration, and force applied to the skin obtained by applying a trial signal waveform as the control signal waveform to the forward dynamics model of the control target representing the relationship with the waveform And a target response waveform parameter corresponding to at least one of the position, velocity, acceleration, and force applied to the skin. Adjust the trial signal waveform based on an error between over down to obtain the target control signal waveform for controlling the asymmetric motion, control signal estimation method. A dynamic characteristic model of a pseudo force sensation generator having a moving member that performs a periodic asymmetric motion according to a control signal waveform, and a gripping portion to which a force generated by the asymmetrical motion is applied, and a contact with the gripping portion Between the control signal waveform and a waveform corresponding to at least one of the position, velocity, acceleration, and force applied to the skin, obtained based on a plurality of models including a skin dynamic characteristic model In the case where an input control signal waveform is applied as the control signal waveform to the forward dynamics model of the control target representing the position, the position, velocity, acceleration, and skin of the gripping part when the asymmetric motion is controlled by the input control signal waveform A waveform estimation method for estimating a response waveform corresponding to at least one of the forces applied to the.   A program for causing a computer to function as the control signal estimation apparatus according to claim 1.   A program for causing a computer to function as the waveform estimation apparatus according to claim 16.