JP7796485B2 - Electric brake device - Google Patents

Description

The present invention relates to an electric brake device that drives an electric motor to control the braking force generated by contact between a friction material and a brake rotor.

Known electric brake devices include those described in the following Patent Documents 1 to 3.

JP 2010-270788 A JP 2012-057681 A Japanese Patent Application Laid-Open No. 2008-184023

When the electric brake device of Patent Document 1 is applied to vehicle brakes, in particular, there is often a need for high-speed, precise control of braking force to avoid causing discomfort to the driver or to improve control performance such as anti-skid control and vehicle motion control.

Generally, in brake devices that use friction material to brake the brake rotor, the stiffness of the brake device, i.e., the change in braking force relative to the amount of motor rotation, is nonlinear with respect to braking force, mainly due to the nonlinearity of the stiffness of the friction material. For this reason, when performing precise control, it is generally preferable to grasp the stiffness as accurately as possible and use information based on that stiffness for control.

However, particularly in vehicle brakes, friction materials generally wear due to repeated braking. As a result, the stiffness of the brake device can change relatively significantly. Furthermore, wear does not progress evenly (uneven wear) due to uneven contact pressure on the friction material, mainly caused by deformation of the brake device. For this reason, stiffness changes depending on the degree of uneven wear, i.e., the unevenness of wear. If this causes a discrepancy between the brake device stiffness stored in the control device and the actual device, there is a risk that the controllability of the braking force of the electric brake device will deteriorate.

Furthermore, when using an actuator that has the function of changing speed using the reaction force of the actuator load, as in Patent Document 2, it is preferable to at least determine in advance the load conditions under which the speed change occurs and to perform control based on the actuator stiffness that reflects the effects of the speed change operation. However, factors such as wear on component parts can change the load conditions under which the speed change occurs, which can lead to problems such as a deterioration in the controllability of the braking force of the electric brake device.

Patent Document 3 also discloses a method for estimating the stiffness of a brake device. However, brake device stiffness generally exhibits relatively strong nonlinearity. For example, it is often difficult to back-calculate using a predetermined mathematical formula from data obtained by actually operating an electric brake device. Instead, it becomes necessary to determine conditions for minimizing error using an iterative method such as Newton's method. However, even in this case, determining a mathematical formula for deriving nonlinear stiffness requires a complex mathematical formula with a large number of variables, resulting in an extremely large computational load and the need for high-performance computing devices, which creates cost issues. Furthermore, as the mathematical formula becomes more complex, the convergence calculation used for estimation may diverge, or the presence of many local solutions may mean that it takes an extremely long time to obtain estimation results with sufficient accuracy.

The object of the present invention is to provide an electric brake device that can estimate stiffness with a small computational load and maintain brake control accuracy even when stiffness changes.

The electric brake device of the present invention comprises a brake rotor, a friction material that contacts the brake rotor to generate a braking force, a friction material operation means that controls the contact state between the friction material and the brake rotor, an electric motor that powers the friction material operation means, and a control device that drives the electric motor to control the braking force generated by the contact between the friction material and the brake rotor. The control device includes an angle estimator that estimates a physical quantity corresponding to the rotation angle of the electric motor, a braking force estimator that estimates braking force, a stiffness estimator that stores the estimated angle estimated by the angle estimator and the estimated braking force estimated by the braking force estimator as an electric brake operation history, and estimates stiffness, which is a correlation between the electric motor rotation amount and braking force, based on the stored estimated angle and estimated braking force, and a braking force control unit that uses the estimated stiffness estimated by the stiffness estimator to derive a motor drive amount to cause the estimated braking force to track a target value for braking force. The stiffness estimator has a stiffness estimation calculation unit that calculates an estimated stiffness using multiple pre-stored reference stiffnesses and a decision variable that determines one of these reference stiffnesses or an intermediate value between them. The stiffness estimation calculation unit compares the calculated value derived from one of the stored estimated angle and estimated braking force using the estimated stiffness with the other stored data to derive an error, and estimates stiffness by performing a convergence calculation that adjusts the decision variable so that the error is reduced to at least a predetermined allowable amount. Here, the "physical quantity equivalent to the rotational angle of the electric motor" is, for example, the stroke amount of the friction material, angular velocity, etc. Angular velocity can be integrated to obtain an angle.

With this configuration, by estimating the stiffness of the electric brake device and performing control calculations based on the estimated stiffness, brake control accuracy can be maintained even if stiffness changes due to wear of the friction material, etc. Furthermore, by deriving the estimated stiffness of the electric brake device using a pre-stored initial stiffness, a reference stiffness that assumes changes in the stiffness of the electric brake device, and a decision variable that combines these, stiffness can be estimated with a relatively small computational load.

In the present invention, the multiple reference stiffnesses stored in the stiffness estimator may be a data table with multiple addresses each containing a reference stiffness, and the decision variable may be an address indicating the reference destination of the data table. With this configuration, representative examples of variable stiffnesses are prepared in advance as a table, and by changing the reference destination of the table and performing a convergence calculation to minimize the error, the design variable becomes only the address indicating the reference destination of the data table. Therefore, an estimated stiffness with a small error compared to the actual stiffness of the electric brake device can be calculated with a relatively small calculation load.

In the present invention, the multiple reference stiffnesses stored in the stiffness estimator may be at least two or more different patterns of reference stiffness, and the decision variable may be a value multiplied as a combination ratio of the multiple reference stiffnesses. With this configuration, the combination ratio is changed when combining representative examples of stiffness that can be changed in advance at a predetermined ratio, and a convergence calculation to minimize error is performed, so that the only design variable is the combination ratio. Therefore, an estimated stiffness with a small error compared to the actual electric brake device stiffness can be calculated with a relatively small calculation load.

In the present invention, the decision variables in the stiffness estimation calculation unit may include a first decision variable that primarily changes the nonlinearity of stiffness, and a second decision variable that primarily changes the overall stiffness.

In the present invention, the friction material operating means may be a linear motion mechanism equipped with a transmission mechanism in which the rotational motion of the electric motor is converted into linear motion, and the equivalent lead, which is the correlation between the amount of rotation and the amount of linear motion, changes at a predetermined braking force; the multiple reference stiffnesses stored in the stiffness estimator may be reference stiffnesses that include changes in the equivalent lead; and the decision variables may include decision variables that primarily change the braking force conditions under which the change in the equivalent lead occurs.

In this case, the friction material operating means may include a planetary reduction mechanism having a rotary input member and planetary rolling elements arranged coaxially with the rotational axis of the rotary input member and spaced equally in the circumferential direction, which generates a deceleration effect depending on the ratio between the orbital speeds of the rotary input member and the planetary rolling elements, an elastic member that applies a fastening force that rotates the rotary input member and the planetary rolling elements as a unit, and a transmission mechanism in which the fastening force of the elastic member is lost due to a reaction force of the pressing force between the friction material and the brake rotor, thereby generating a planetary reduction effect; the multiple reference stiffnesses stored in the stiffness estimator are reference stiffnesses that include the amount of deformation of the elastic member; and the decision variables in the stiffness estimation calculation unit may include a decision variable that changes the braking force conditions at which deformation of the elastic member is complete.

In the present invention, the stiffness estimator may be configured to determine the amount of braking operation based on at least one of the amount of change in the estimated braking force and the amount of change in the estimated angle in the electric brake operation history, and to limit the change in the decision variable in the convergence calculation when estimating the stiffness when the amount of braking operation becomes small. The smaller the operating range of the electric brake used for error estimation, the greater the risk of an erroneous estimation result due to the influence of measurement noise, fluctuations in braking force (e.g., uneven thickness of the brake disc of a disc brake, etc.). Therefore, for example, by narrowing the range in which the decision variable can be changed relative to the previous estimation result, the risk of deriving an erroneous estimation result can be avoided.

In the present invention, the angle estimator may estimate the angular velocity of the brake rotor, and the stiffness estimator may determine the degree of brake use based on at least one of the time during which the estimated angle and estimated brake force from the electric brake operation history are acquired, the angular velocity of the brake rotor, and the estimated brake force, and may be configured to limit changes in the decision variable in the convergence calculation when estimating the stiffness when the degree of brake use decreases. The longer a friction material is pressed against a brake rotor rotating at high speed with a greater force, the more likely the friction material is to wear, and in the opposite situation, it is considered to be relatively resistant to wear. Therefore, for example, by narrowing the range within which the decision variable can be changed relative to the previous estimation result, the risk of deriving an erroneous estimation result can be avoided.

In the present invention, the control device may have a driving state estimator that estimates the driving state of the vehicle equipped with the electric brake device, and the stiffness estimator may have a function to determine whether stiffness estimation has been performed based on an electric brake operation history in which a change in estimated brake force greater than a predetermined value has occurred, and to measure a non-execution time during which stiffness estimation is not performed based on the determination, and a function to estimate, when the non-execution time has elapsed or more, that the vehicle equipped with the electric brake device has been stopped for a predetermined time or more as the driving state, and, if the vehicle's braking force is smaller than a predetermined value, to generate a braking force greater than the predetermined braking force regardless of operation by the driver of the vehicle, and perform stiffness estimation using the estimated braking force and estimated angle at that time.

To make a highly reliable stiffness estimate, it is preferable to use data from situations where a sufficiently large braking force is generated. However, it is conceivable that situations may arise where a braking force equivalent to or greater than the normal braking range (generally a braking range of less than approximately 0.2 G) is not generated for an extended period of time. For this reason, in such situations, braking force for estimating stiffness can be automatically generated and the stiffness estimated while the vehicle is stopped, where there is no problem if the braking force is at least greater than required, thereby obtaining an estimated stiffness that is in line with the actual vehicle conditions.

The electric brake device of the present invention allows stiffness to be estimated with a small computational load, and brake control accuracy can be maintained even if stiffness changes.

1 is a configuration diagram of an electric brake device according to a first embodiment of the present invention; FIG. 3 is a configuration diagram of an electric brake device according to a first modified example of the first embodiment. FIG. 10 is a configuration diagram of an electric brake device according to a second modified example of the first embodiment. FIG. 10 is a configuration diagram of an electric brake device according to a third modified example of the first embodiment. FIG. 2 is a diagram illustrating a brake system configured using a plurality of the electric brake devices. FIG. 2 is a configuration diagram of another example of the brake system. 2 is a configuration diagram showing an example of a brake controller and a stiffness estimator of the electric brake device. FIG. FIG. 4 is a configuration diagram showing another example of the brake controller and the stiffness estimator of the same. FIG. 2 is a block diagram showing an example of reference stiffness and decision variables in a stiffness estimation calculation unit of the stiffness estimator. FIG. 10 is a block diagram illustrating another example of the reference stiffness and the decision variables. FIG. 10 is a block diagram illustrating yet another example of the reference stiffness and the decision variables. FIG. 10 is a block diagram illustrating yet another example of the reference stiffness and the decision variables. FIG. 10 is a block diagram illustrating yet another example of the reference stiffness and the decision variables. 10 shows yet another example of the same reference stiffness and the same decision variables. FIG. 10 is a flow diagram showing an example of a stiffness estimation flow. FIG. 10 is a flowchart showing an example of a flow for storing an estimated angle and an estimated braking force in an operation history storage unit of the stiffness estimator. FIG. 10 is a flow diagram showing an example of a flow for setting different limit ranges for decision variables during stiffness estimation. FIG. 10 is a flow diagram showing another example of a flow for setting different limit ranges for decision variables during stiffness estimation. FIG. 10 is a flow diagram showing an example of a flow for operating the electric brake device under conditions that allow automatic stiffness estimation and performing stiffness estimation when the conditions for performing stiffness estimation are not satisfied for a long period of time. 1A is a graph showing an example of the operation of the electric brake device, and FIG. 1B is a graph showing an example of the operation of a conventional electric brake device.

Figure 1 shows the configuration of an electric brake device 1 according to a first embodiment of the present invention. The electric brake device 1 includes a control device 2, a brake actuator 6 using a linear motion mechanism 4, and a brake command means 8 such as a brake pedal.

[Structure of Brake Actuator]
The brake actuator 6 has an electric motor 10, a linear motion mechanism 4, an angle sensor 12, a load sensor 14, and a reducer 16. The angle sensor 12 detects the motor angle of the electric motor. The load sensor 14 detects the brake load of the electric brake device 1. Note that the reducer 16 may be omitted depending on the performance requirements of the brake.

The electric motor 10 is, for example, a permanent magnet synchronous motor. Using a permanent magnet synchronous motor as the electric motor 10 saves space, and provides high efficiency and high torque. However, the electric motor 10 is not limited to a permanent magnet synchronous motor, and may be, for example, a DC motor using brushes, a reluctance motor without permanent magnets, an induction motor, etc. Furthermore, the electric motor 10 may be a radial gap motor with magnetic poles arranged in the direction of the rotational radius, or an axial gap motor with magnetic poles arranged in the direction of the rotational axis.

The linear motion mechanism 4 can be a screw mechanism such as a planetary roller screw, ball screw, or sliding screw, or a ball ramp mechanism or other mechanism capable of converting rotational motion into linear motion.

The angle sensor 12 is, for example, a resolver or a magnetic encoder. Using a resolver, magnetic encoder, or the like as the angle sensor 12 provides high accuracy and reliability. However, the angle sensor 12 is not limited to these, and various sensors such as an optical encoder can be used. Alternatively, the motor angle may be estimated without using the angle sensor 12 (angle sensorless estimation), for example, from the relationship between voltage and current.

The load sensor 14 is, for example, a sensor that detects strain, deformation, etc. corresponding to the load acting on the actuator 6. Using such a sensor is inexpensive and highly accurate. However, the load sensor 14 is not limited to this and may also be a pressure-sensitive medium such as a piezoelectric element. Alternatively, the load sensor 14 may be a torque sensor that detects the braking torque of a brake rotor, or, in the case of an electric vehicle brake system, an acceleration sensor that detects the longitudinal deceleration of the vehicle. Furthermore, without using the load sensor 14, for example, the torque generated by the motor may be estimated from the motor current, and the braking force may be estimated (load sensorless estimation) using the characteristics of a linear motion mechanism, etc., from the motor torque and the motor's motion state.

Although not shown, various sensors such as thermistors may be installed separately as needed. Furthermore, a mechanism for locking the actuator's power transmission section using a solenoid, DC motor, etc. may be installed, allowing it to be used as a parking brake actuator.

[Configuration of the control device]
The control device 2 includes a brake controller 18 , a motion state estimator 20 , a braking force estimator 22 , a stiffness estimator 24 , a motor controller 26 , a motor driver 28 , and a current sensor 30 .

The brake controller 18 performs calculations for brake control. The motion state estimator 20 calculates the operating state of the electric motor 10. The brake force estimator 22 estimates the brake force. The stiffness estimator 24 estimates the stiffness of the electric brake device 1. The motor controller 26 controls the motor current to obtain a predetermined motor output. The motor driver 28 supplies power to the electric motor 10. The current sensor 30 detects the motor current supplied to the electric motor 10.

The motion state estimator 20 has an angle estimator 32 that estimates the rotor angle of the electric motor 10, and an angular velocity estimator 34 that estimates its angular velocity. In addition, the motion state estimator 20 may also have functions such as estimating predetermined differential and integral values such as the angular acceleration of the electric motor 10, and estimating disturbances.

The angle of the electric motor 10 includes physical quantities required based on the control configuration, such as the electrical angle phase used for current control and the total rotation angle corrected for overlap and underlap of the angle sensor 12 used for angle control. Furthermore, the angle and angular velocity of the electric motor 10 may be, instead of the rotor of the electric motor 10, for example, the angle of a specific part of the reducer calculated based on the reduction ratio, or the position and velocity calculated based on the equivalent lead of a screw mechanism. Estimation of such physical quantities may use, for example, a configuration such as a state estimation observer, or may be direct calculation such as back calculation based on differentiation or inertia equations.

The current sensor 30 may be, for example, a sensor consisting of an amplifier that detects the voltage across a shunt resistor installed in the current path, or a non-contact sensor that detects magnetic flux around the current path. However, the current sensor 30 is not limited to these, and may also be one that detects the terminal voltage of elements that make up the motor driver 28. The current sensor 30 may also be installed between phases of the electric motor 10, or one or more current sensors may be installed on the low side or high side. Alternatively, feedforward control may be performed based on motor characteristics such as inductance and resistance without using a current sensor.

The brake controller 18 determines the amount of operation required for the brake actuator to follow a predetermined command input from the brake instruction means (brake pedal) 8, and converts this into a motor drive signal. The brake controller 18 has a braking force control unit 36 and a stiffness memory unit 38. The braking force control unit 36 mainly controls the braking force generated by contact between the friction material 40 and the brake rotor 42. The stiffness memory unit 38 stores the stiffness of the electric brake device 1 used in braking force control.

The braking force control unit 36 determines the motor drive amount so that the braking force when the friction material 40 and the brake rotor 42 are brought into contact is controlled to track the desired target value. In FIG. 1, the pressing force between the friction material 40 and the brake rotor 42 is detected by the load sensor 14, and the motor drive amount is determined based on the braking force estimated by the braking force estimator 22 from the output of this load sensor 14. However, the method for determining the motor drive amount is not limited to this, and another sensor such as a torque sensor that detects the braking torque of the brake rotor 42 as described above may be used, or the estimated braking force may be feedback-controlled using a load sensorless estimation function.

The braking force control unit 36 may have a position control function that primarily controls the stroke position of the linear motion mechanism 4. The position control function determines the motor drive amount to control the linear motion stroke amount of the linear motion mechanism 4, which is converted from the motor rotation amount based on the actuator specifications, such as the equivalent lead when a screw mechanism is used or the reduction ratio when a reducer is provided. In this case, there is no need to provide a separate stroke sensor, etc. However, a separate stroke sensor, etc. may also be provided, and the signal from the sensor may be feedback-controlled to a predetermined target value.

The position control function can, for example, control the stroke amount to ensure a desired gap exists between the friction material 40 and the brake rotor 42, so that they do not come into contact as much as possible when the brake is released. Additionally, in order to control extremely slight braking forces that are difficult to detect using load sensors or torque sensors that detect braking force, the brake force control function can also control the stroke so that the gap is close to or less than zero. In other words, the linear motion mechanism 4 constitutes a friction material operating means that operates the contact state between the friction material 40 and the brake rotor 42.

The stiffness memory unit 38 stores the stiffness of the electric brake device 1, which is the correlation between the motor rotation amount and the brake load. The stiffness memory unit 38 stores the stiffness of the electric brake device 1, which has been determined in advance, for example, through experiments or analysis, as an initial condition, and the stored stiffness is updated based on the estimation results of the stiffness estimator 24.

The stiffness estimator 24 has an operation history storage unit 44 and a stiffness estimation calculation unit 46. The operation history storage unit 44 stores the estimated braking force and estimated angle of the electric motor 10 when generating braking force as the electric brake operation history. The stiffness estimation calculation unit 46 estimates the stiffness of the electric brake device 1 from the estimated braking force and estimated angle stored in the operation history storage unit 44.

The stiffness estimation calculation unit 46 has multiple preset reference stiffnesses and decision variables that can uniquely determine stiffness from these reference stiffnesses based on a predetermined calculation process. The stiffness estimation calculation unit 46 derives stiffness based on these reference stiffnesses and decision variables. The reference stiffness is determined in advance, for example, through experiments or analysis, and stored in the stiffness estimation calculation unit 46. The stiffness estimation calculation unit 46 further estimates stiffness based on data stored in the operation history storage unit 44. Specifically, the stiffness estimation calculation unit 46 evaluates the error between the stiffness derived by changing the decision variables and the stored data, and derives an estimated stiffness through convergence calculation that minimizes the error or reduces the error to a predetermined value.

Generally, the stiffness of the electric brake device 1 exhibits extremely strong nonlinearity, mainly due to the influence of the friction material 40. For this reason, it is often difficult to derive complex approximation formulas using relatively inexpensive microprocessors, etc. In this embodiment, stiffness information is provided in advance as reference stiffness, and the decision variable for uniquely deriving stiffness from multiple reference stiffnesses is used as the search parameter in the convergence calculation for error minimization, thereby significantly reducing the calculation load.

Furthermore, for example, when operation is required based on an arbitrary operation by a specified driver, as typified by the electric brake device 1 mounted on a vehicle, the operation will not necessarily generate braking force from minimum to maximum. Therefore, estimation must be made based on data from limited electric brake operation. Even in such cases, according to this embodiment, stiffness is estimated based on a specified reference stiffness, thereby reducing the risk of requiring an estimated stiffness that is significantly different from that of the actual device outside the range of data.

The motor controller 26 controls the motor current to achieve the desired motor drive amount determined by the brake controller 18. For example, the optimal current conditions for obtaining the desired torque at a given motor angular velocity are stored in a look-up table (LUT) in advance, and the motor current is controlled to achieve a target current value determined from the current motor angular velocity. In this case, highly accurate control can be achieved inexpensively. However, drive conditions can also be determined in real time by calculating the relationship between current and voltage that derives motor output.

The above various calculation functions are ideally implemented using calculators and peripheral circuits, such as microcomputers, FPGAs, and ASICs, as this provides low cost and high performance.

The motor driver 28 forms a bridge circuit using switching elements such as FETs, and performs PWM control, which determines the voltage applied to the motor using a predetermined duty ratio. This configuration results in low cost and high performance. However, the motor driver 28 may also be configured to perform PAM control by providing a transformer circuit or the like.

[others]
The electric brake device 1 further includes a power supply device (not shown). In an electric brake device for an automobile, for example, the power supply device may be a low-voltage battery or a step-down converter that steps down a high-voltage battery. Alternatively, the power supply device may use a high-capacity capacitor or the like, or these may be used in parallel to provide redundancy. Although not shown, it is preferable to supply power directly to the motor driver 28 and use a small step-down converter in the control device 2 for the computing unit, etc., but it is also possible to supply power to the motor driver 28 via a step-up converter.

Instead of a brake pedal, various control means operable by the driver, such as a volume, joystick, switch, etc., may be used as the brake command means 8.

Figure 2 shows a first variant in which a vehicle motion control device 48 is provided as a brake command means. The vehicle motion control device 48 has, for example, an automatic braking function unit, a skid prevention function unit, an anti-skid control unit, etc. (none of which are shown). The automatic braking function unit prevents vehicle collisions or reduces the impact of a collision. The skid prevention function unit prevents the vehicle from spinning out by at least braking when the vehicle enters a skid state. The anti-skid control unit prevents the wheels from locking due to braking, resulting in unstable vehicle behavior.

The vehicle motion control device 48 may be an integrated control device that integrates information from various on-board sensors (none of which are shown), such as a G sensor, object sensor, and GPS, and performs the calculations required for the various functions described above. The brake operation amount determined by the vehicle motion control device 48 is transmitted to the electric brake control device 2 as the target brake force. Figure 3 shows a second modified example in which command transmission to the brake control device 2 is integrated into the vehicle motion control device 48.

Figure 4 shows a third variation of Figure 1, in which an actuator is used in the linear motion mechanism 4, in which the correlation between the equivalent lead (= the amount of linear motion of the actuator relative to the amount of motor rotation) changes under specified operating conditions.

The linear motion mechanism 4 in Figure 4 has a speed change mechanism 50 whose equivalent lead changes depending on the linear load applied. The linear motion mechanism 4 is, for example, a planetary roller screw structure in which the equivalent lead changes as the planetary carrier and planetary rolling elements are fastened and separated by the reaction force of the linear load. Using such a planetary roller screw structure simplifies the configuration and saves space. However, the linear motion mechanism 4 is not limited to a planetary roller screw structure. For example, a linear motion mechanism that combines a planetary reducer with the speed change mechanism described above with various mechanisms capable of converting rotational motion into linear motion, such as a ball screw or ball ramp mechanism, may also be used.

In a linear motion mechanism 4 equipped with a transmission mechanism 50 using the aforementioned planetary reduction structure, when the planetary carrier and planetary rolling elements are fastened and rotate together, the equivalent lead is relatively large. On the other hand, when the planetary carrier and planetary rolling elements separate and the planetary rolling elements rotate, the equivalent lead is relatively small due to the planetary reduction effect. When such an actuator is used, for example, when the brake is released and braking force is generated from a state where a predetermined clearance is provided between the friction material 40 and the brake rotor 42, the equivalent lead is large when the load is light, resulting in fast operation. As the load increases, the equivalent lead decreases, allowing a larger thrust to be generated, thereby improving brake responsiveness.

When a gear shift occurs, the stiffness of the electric brake device 1 becomes discontinuous depending at least on the presence or absence of the planetary deceleration effect. In other words, the gradient of change in either the motor rotation amount or the brake load relative to the other becomes discontinuous at least by the deceleration ratio due to the planetary deceleration effect. In addition, for example, if a spring member that deforms due to the reaction force of the linear load is used as a means for fastening and separating the planet carrier and planetary rolling elements described above, it is preferable from a performance perspective that the deformation amount of the spring member is as close to zero as possible, but due to manufacturing considerations such as dimensional tolerances, the structure requires that the spring member deform by a predetermined amount. Therefore, while the spring member deforms, the stiffness of the spring member affects the actuator stiffness, resulting in discontinuous stiffness.

The stiffness estimator 24 stores stiffness information that assumes discontinuous stiffness as reference stiffness. Alternatively, the stiffness estimator 24 is provided with a reference stiffness that does not have discontinuities, and a decision variable that adds discontinuities associated with the transmission structure 50 to the reference stiffness.

Figure 5 shows an example of a brake system made up of multiple electric brake devices 1 with any of the configurations shown in Figures 1 to 4. In Figure 5, the target brake force generated by the vehicle motion control device 48 is transmitted independently to each electric brake device 1. Furthermore, while the brake pedal 8 has a single piece of information, such as the pedal stroke amount, the actual brake force generated for the same brake pedal 8 stroke amount can be different for each electric brake control device 1, such as the front and rear brakes of a four-wheeled vehicle. The specifications and the number of electric brake devices 1 configured as a system can be determined appropriately depending on the requirements of the brake system.

Figure 6 shows another example of a brake system, different from Figure 5, that is configured with multiple electric brake devices 1 having any of the configurations shown in Figures 1 to 4. Figure 6 shows an example in which one electric brake device 1 controls multiple brake actuators 6. While Figure 6 shows an example in which two brake actuators are controlled by one electric brake control device 1, the number of brake actuators 6 controlled by one electric brake control device 1 can be determined appropriately depending on the system requirements. The configurations shown in Figures 5 and 6 may also be used together.

In the examples of Figures 1 to 6, the illustrated functional blocks are provided solely for the convenience of explanation and do not restrict the hardware or software configuration or partitioning. Furthermore, the specific software and hardware configuration can be configured as desired as long as it does not interfere with the illustrated functions, and the functions of each illustrated block may be integrated or divided as needed. Furthermore, elements not illustrated may be added as long as it does not interfere with the illustrated functions. For example, safety mechanisms for when various functions or sensors fail may be added as appropriate based on the system requirements.

The illustrated electric brake device 1 can be used as a brake device to stop not only automobiles, but also elevators, power generators, energy storage devices such as flywheels, etc.

Figure 7 shows an example configuration of the brake controller 18 and stiffness estimator 24. The position control unit 54, braking force control unit 36, angle conversion unit 56, and stiffness memory unit 38 make up the brake control unit 18 in Figure 1.

The position control unit 54 derives the motor drive amount for tracking and controlling the estimated angle relative to the target motor angle value. Here, the angle in Figure 7 does not refer to an overlapping/underlapping angle with a predetermined periodicity, such as an electrical angle or mechanical angle, but rather refers to the total rotation angle indicating the total amount of rotation of the electric motor. Also, while Figure 7 shows an example in which the motor drive amount is the motor torque target value, it may also be, for example, a motor current target value or motor voltage.

The braking force control unit 36 compares the motor angle converted from the estimated braking force based on the estimated stiffness with the estimated angle, and calculates an angle compensation value that results in the desired braking force. In other words, the braking force control unit 36 compensates for the error between the actual stiffness of the electric brake device 1 and the estimated stiffness. The angle compensation value may be applied to the target value in the position control unit 54, or may be applied to the fed-back estimated angle.

The angle conversion unit 56 converts the braking force into a motor angle based on the stiffness stored in the stiffness memory unit 38.

The stiffness memory unit 38 stores the stiffness of the electric brake device 1, which is the correlation between braking force and motor angle. The stiffness memory unit 38 pre-stores an initial stiffness value that corresponds to the initial state of the electric brake device 1. Thereafter, when a new stiffness is estimated by the stiffness estimation calculation unit 46, it is updated to that estimated stiffness.

The operation history storage unit 44 has a braking force storage unit 58 and an angle storage unit 60. The braking force storage unit 58 stores the estimated braking force when braking force is generated by the electric brake device 1. The angle storage unit 60 stores the estimated angle when braking force is generated by the electric brake device 1.

The stiffness estimation calculation unit 46 has a plurality of reference stiffnesses 62 given in advance, a decision variable 64, a stiffness calculation unit 66, and a convergence calculation unit 68. The stiffness calculation unit 66 derives stiffness from the reference stiffness 62 and the decision variable 64. The convergence calculation unit 68 derives a decision variable 64 that minimizes the error with the data stored in the operation history storage unit 44 or makes the error smaller than a predetermined value.

The reference stiffness 62 includes at least the stiffness of the electric brake device 1 in its initial state and the stiffness of the electric brake device 1 assuming that the stiffness of the electric brake device 1 has changed. Multiple reference stiffnesses 62 may be provided for each assumed change trend, assuming different trends in stiffness changes. For example, stiffness changes caused by uneven wear of the brake's friction material 40 and stiffness changes caused by the friction material 40 becoming thinner due to wear generally have different trends. For this reason, multiple reference stiffnesses 62 may be provided, including a reference stiffness 62 assuming stiffness changes caused by uneven thickness of the friction material 40 and a reference stiffness 62 assuming stiffness changes caused by the friction material 40 becoming thinner overall, and multiple independently adjustable decision variables 64 may be provided for each. It is also possible to add reference stiffnesses 62 assuming stiffness changes that show more trends, and independently adjustable decision variables 64 for each.

The convergence calculation unit 68 uses a convergence calculation algorithm, such as Newton's method, for the data samples θ1...θk and F1...Fk (k is the number of samples) in the operation history storage unit 44 to derive a decision variable 64 that minimizes the next error function J or makes the error smaller than a predetermined value.

However, the error function J does not have to be based on the squared error described above, but can be any calculation formula that can evaluate the error between the result derived using stiffness and the actual data in the operation history memory unit 44, such as the sum of absolute values of differences, Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), etc.

Furthermore, in addition to minimizing the error, it is preferable to set a maximum number of iterations for the convergence calculation, and to provide a process to terminate the calculation when the number of iterations in the convergence calculation reaches the maximum number of iterations. By providing such a process, it is possible to prevent situations in which extremely long iterative calculations are performed. In this case, for the estimation result when the termination condition of reaching the maximum number of iterations is met, the estimated stiffness may be discarded, or it may be updated to the calculated stiffness with the smallest error. Alternatively, the error under the minimized conditions may be evaluated, and if such error is smaller than a predetermined value, the stiffness in the stiffness memory unit 38 may be updated to the estimated stiffness, or if the error is larger, the result may be discarded.

Although not shown, when updating the stiffness of the stiffness memory unit 38, the updated stiffness may be saved in a separate memory area from the stiffness before the update. This reliably prevents operational malfunctions such as referencing stiffness information that is in the middle of being updated when the angle conversion unit 56 references the stiffness memory unit 38. In this case, for example, if a memory area capable of storing two stiffnesses is provided and one is updated and then the other is updated, and the two areas are used alternately, the amount of memory area used can be minimized. However, for example, an area capable of storing more than one stiffness may be provided and multiple updates' worth of information may be held. Completion of the stiffness update can be managed, for example, by providing a flag process indicating that the update has been completed.

Although not shown, the operation history memory unit 44 and stiffness estimation calculation unit 46 may be configured to include a first operation history memory unit that stores only data from operations when the braking force increases, a first stiffness estimation calculation unit that performs calculations using data from this first operation history memory unit, a second operation history memory unit that stores only data from operations when the braking force decreases, and a second stiffness estimation calculation unit that performs calculations using data from this second operation history memory unit, thereby performing stiffness estimation when the braking force increases and decreases.

This configuration allows for more accurate stiffness estimation, for example, in cases where factors such as hysteresis due to frictional force cause different stiffnesses when the braking force increases and decreases. In such cases, the stiffness storage unit 38 may store either the estimated stiffness when the braking force increases or when the braking force decreases, or an intermediate value between the two. Alternatively, the estimated stiffnesses when the braking force increases and decreases may be stored, and the estimated stiffness when the braking force increases may be used when the braking force increases, and the estimated stiffness when the braking force decreases may be used when the braking force decreases.

Fig. 8 shows an example of the brake controller 18 and stiffness estimator 24 that is different from that shown in Fig. 7. The brake force control section 36, stiffness storage section 38 and control gain calculation section 70 constitute the brake control section 18 of Fig. 1 .

The braking force control unit 36 derives the motor drive amount so that the estimated braking force follows the braking force target value. While Figure 8 shows an example in which the motor drive amount is the motor torque target value, the motor drive amount may also be, for example, the motor current target value or the motor voltage.

The control gain calculation unit 70 derives the control calculation parameters for the braking force control unit 36 based on the stiffness stored in the stiffness memory unit 38. The control calculation parameters based on the stored stiffness can be determined in advance through analysis or experimentation as control parameters that maintain constant control characteristics against changes in spring rate, for example, when the controlled object is a spring-coupled mass and changes in stiffness are considered to be changes in spring rate. Furthermore, the control calculation parameters may be, for example, the gains of a serial linear compensator such as PID control, the coefficients of a state feedback controller, or other predetermined parameters regardless of the controller format.

Figure 9 shows an example of the reference stiffness 62 and decision variable 64 in the stiffness estimation calculation unit 46 shown in Figures 7 and 8. In the example of Figure 9, the reference stiffness 62 is a stiffness table (data table) 72 containing multiple pieces of stiffness information, and the decision variable 64 is an address that specifies the reference destination of the stiffness table 72.

Figure 9 shows an example of a two-dimensional table with first and second addresses. In this case, the reference address is an address containing two values specifying the first and second addresses. For example, in the example of Figure 9, the first address specifies a reference stiffness that assumes stiffness changes due to uneven thickness of the friction material 40 (Figure 1), and the second address specifies a reference stiffness that assumes stiffness changes due to thinner friction material 40. This allows for a reference stiffness 62 and decision variable 64 that take into account stiffness changes due to different trends.

Specifically, in Figure 9, stiffness f_00 is the reference stiffness in the initial state. Stiffness f_0m is the reference stiffness when the average thickness of the friction material 40 is at its most uneven, roughly at its initial state. Stiffness f_n0 is the reference stiffness when the unevenness of the friction material 40 is at its thinnest average thickness, roughly at its initial state. Stiffness f_nm is the reference stiffness when the thickness of the friction material 40 is at its most uneven and the thinnest average thickness. If a stiffness table 72 is provided with reference stiffnesses 62 that are intermediate between addresses 0 to m and 0 to n, it is possible to obtain a reference stiffness that takes into account stiffness changes due to both unevenness and average thickness of the friction material 40. This makes it possible to change the degree of stiffness change using the reference address.

Note that the above setting example is merely an example, and the number of dimensions of the table and address can be set arbitrarily. Furthermore, the type of stiffness change factor assumed for each dimension of the table can also be set arbitrarily. That is, it may be a one-dimensional table and address, or a three- or more-dimensional table and address. Furthermore, the reference address may directly specify the address of the table, or it may be possible to specify the middle address of the table. For example, if it is possible to specify the middle address in a two-dimensional reference table, the reference stiffnesses at the four addresses surrounding the reference address may be linearly interpolated to obtain the combined stiffness.

Figure 10 shows another example of the reference stiffness 62 and decision variables 64 in the stiffness estimation calculation unit 46 shown in Figures 7 and 8, different from those shown in Figure 9. Figure 10 shows an example in which multiple reference stiffnesses 62 and coupling coefficients α and β that couple these reference stiffnesses 62 are used as decision variables 64.

Figure 10 shows an example consisting of two coupling coefficients α and β and first to fourth reference stiffnesses 62 coupled by these. For example, in Figure 10, the first reference stiffness is the stiffness in the initial state. The second reference stiffness is the stiffness in a state where the average thickness of the friction material 40 is at its most uneven, roughly at its initial state. The third reference stiffness is the stiffness in a state where the unevenness of the friction material 40 is at its thinnest average thickness, roughly at its initial state. The fourth reference stiffness is the stiffness in a state where the thickness of the friction material 40 is at its most uneven and where the average thickness is at its thinnest. In this case, the reference stiffness can be determined by taking into account the stiffness changes due to the unevenness and average thickness of the friction material 40, and the degree of stiffness change can be changed by the coupling coefficients α and β.

Note that the above setting example is merely an example, and the number of coupling coefficients and the associated number of reference stiffnesses can be set arbitrarily. Furthermore, the stiffness change factors assumed for each reference stiffness can be set arbitrarily. That is, there can be one coupling coefficient and two reference stiffnesses, or three coupling coefficients and eight reference stiffnesses, or even more groups of coupling coefficients and reference stiffnesses.

Figure 11 shows an example in which one coupling coefficient α, two reference stiffnesses 62, and a multiplication coefficient ε that changes the overall stiffness are provided. For example, when the electric brake device 1 is configured based on specifications that primarily require a relatively small maximum brake load, the change in stiffness when the average thickness of the friction material 40 decreases can sometimes be approximated as a roughly geometric change in the overall stiffness. In such cases, as shown in Figure 11, stiffness changes that exhibit complex nonlinearity, such as uneven thickness of the friction material 40, are calculated using the first and second reference tables and the coupling coefficient α, and the overall stiffness is changed geometrically using the multiplication coefficient ε. This makes it possible to perform stiffness estimation that adapts to stiffness changes that exhibit different trends, such as the effects of changes in unevenness of the friction material 40 and changes in the average thickness of the friction material 40.

Figure 12 shows an example of a configuration in which a coupling coefficient α and corresponding first and second reference stiffnesses 62, and a coupling coefficient κ and corresponding first and second adjustment functions 74 are provided, and the stiffness is finally obtained by multiplication. Generally, the change in stiffness when the average thickness of the friction material 40 becomes thinner tends to change at a relatively large rate in areas where the braking force is low and at a relatively small rate in areas where the braking force is high. This stiffness change rate is specified in advance as the adjustment function 74, and the degree to which the adjustment function 74 is applied is adjusted using the coupling coefficient κ. This makes it possible to perform stiffness estimation that adapts to stiffness changes that exhibit different trends, such as the influence of changes in the unevenness of the friction material 40 and the influence of changes in the average thickness of the friction material 40.

When using a linear motion mechanism 4 equipped with a transmission structure 50 such as that shown in Figure 4, the stiffness of the electric brake device 1 becomes an even more complex function, due to factors such as the shifting operation and deformation of the spring members that make up the transmission mechanism 50. In this case, it becomes more difficult to estimate stiffness while taking into account changes in the characteristics of the shifting operation (for example, changes in braking force that cause shifting, spring deformation characteristics, etc.). Even in such cases, stiffness can be estimated using the methods shown in Figures 9 to 12.

For example, when applying the example of FIG. 9, the reference stiffness table 72 can be made three-dimensional, and a reference stiffness 62 that takes into account changes in the characteristics of the gear shifting operation can be mapped to the third address. Furthermore, when applying the example of FIG. 10, a new coupling coefficient and fifth to eighth reference stiffnesses 62 can be set, and the fifth to eighth reference stiffnesses 62 can be reference stiffnesses that reflect changes in the characteristics of the gear shifting operation in addition to the first to fourth reference stiffnesses 62. Furthermore, when applying the examples of FIG. 11 and FIG. 12, a new coupling coefficient and third and fourth reference stiffnesses 62 can be set, and the third and fourth reference stiffnesses 62 can be reference stiffnesses that reflect changes in the characteristics of the gear shifting operation in addition to the first and second reference stiffnesses 62.

Figure 13 shows an example of directly calculating the characteristics of a gear shift operation instead of the method described above. First, as in Figure 9, a predetermined reference stiffness is derived using the reference stiffness table 72 and reference address. Next, using the discontinuity point FV1 where a gear shift occurs, the discontinuity point FV2 where spring deformation reaches its limit, and a predetermined calculation, the discontinuity calculation unit 76 derives stiffness including the discontinuity due to the gear shift operation based on the reference stiffness. At this time, the reference address and discontinuity points FV1 and FV2 become decision variables 64, and the stiffness estimator 24 derives decision variables that minimize error in its convergence calculation.

In the discontinuous point calculation unit 76, when the reference stiffness is calculated based on the reference stiffness table 72 in a state where the equivalent lead of the linear motion mechanism 4 (FIG. 4) is small, the following equation is obtained for the reference stiffness θ and F when the braking force is between zero and FV1:
dθ'/dF = R dθ/dF (R is the reduction ratio due to gear change)
This can be said as a calculation that reflects the fact that, for braking forces smaller than the braking force FV1 at which a discontinuity occurs due to a gear shift, the equivalent lead increases with respect to the reference stiffness 62 due to the gear shift.

When the braking force is between FV1 and FV2,
dθ'/dF=(dθ/dF)+v
This can be said as a calculation that reflects the fact that the rigidity of the electric brake device decreases by an amount corresponding to the deformation of the spring members of the speed change structure until the deformation of the spring members reaches its limit.

In the region where the braking force exceeds FV2, dθ'/dF = dθ/dF. In other words, by combining the above θ', the correlation between θ' and F becomes the final derived stiffness.

In the discontinuous point calculation unit 76, when the reference stiffness is derived based on the reference stiffness table 72 in a state where the equivalent lead of the linear motion mechanism 4 is large, in a range where the braking force exceeds FV1, the reference stiffness θ and F are calculated as follows:
dθ'/dF=S·dθ/dF (where S=1/R)
This can be said as a calculation that reflects the fact that, for braking forces larger than the braking force FV1 at which a discontinuity occurs due to a gear shift, the equivalent lead becomes smaller with respect to the reference stiffness 62 due to the gear shift.

When the braking force is between FV1 and FV2, and further taking into consideration the deformation of the spring member,
dθ'/dF=S・(dθ/dF)+v
Derive θ' such that:

In the region where the braking force is smaller than FV1, dθ'/dF = dθ/dF. In other words, by combining the above θ', the correlation between θ' and F becomes the final derived stiffness.

Note that if the transmission mechanism 50 (Figure 4) does not include a spring member, or if the spring member is designed to have negligible deformation, the calculation steps for FV2 and FV2-related factors described above can be omitted.

Figure 14 shows an example of adding calculation of characteristics due to gear shifting to Figure 10 instead of Figure 9, as compared to Figure 13. The calculation method of the discontinuity calculation unit 76 is the same as that shown in Figure 13.

Figure 15 shows an example of a stiffness estimation flow. Step S1 acquires estimated angle and estimated braking force data from the operation history storage unit 44.

Step 2 determines whether the data acquired in step S1 is sufficient for stiffness estimation. This determination can be made, for example, if one or both of the ranges of change in the estimated braking force and the ranges of change in the estimated angle in the acquired data are greater than predetermined values.

Step S3 sets the initial value of the decision variable 64, which is a variable in the convergence calculation to minimize error when performing stiffness estimation. Setting the initial value to the decision variable 64 that determines the estimated stiffness (stiffness before updating) at the start of the estimation calculation is often preferable, as it improves the convergence of subsequent convergence calculations.

Step S4 represents an iterative loop for convergence calculations. Iterative calculation methods such as Newton's method or the Levenberg-Marquardt method can be used as convergence calculation algorithms. Conditions for terminating such a loop can be set, for example, based on convergence to a minimum error value, the absolute value of the error, or the maximum number of iterations, and can be determined appropriately by the designer in accordance with the algorithm and requirements. If a maximum number of iterations is specified and the iterative calculations end at that number, the calculation results may be discarded as insufficient convergence has been achieved, or the optimal solution among the calculated results may be applied as the result. It is also possible to provide a process for adopting the estimated error in the optimal solution among the calculated results if the error is smaller than a predetermined value, and discarding the result if the error is larger.

Step S5 derives a stiffness function based on the decision variable 64 and reference stiffness 62 in the current iteration loop. In the examples of Figures 9 and 13, the decision variable 64 is a reference address, and the reference stiffness 62 is a reference stiffness table 72 that stores multiple reference stiffnesses. In the examples of Figures 10-12 and 14, the decision variable 64 is coupling coefficients α and β, and the reference stiffness 62 is multiple stored reference stiffnesses. In the example of Figure 11, the multiplication coefficient ε is added to the decision variable 64, and in the examples of Figures 13 and 14, the discontinuity points FV1 and FV2 are added to the decision variable 64 (FV2 may be omitted).

Step 6 evaluates the error between the results derived using the stiffness function determined in step 5 and the actually measured data. This error can be evaluated using the sum of the squared differences as shown in the figure, but evaluation indices such as the sum of the absolute values of the differences, Akaike Information Criterion, or Bayes Information Criterion can also be used.

In step 7, based on the evaluation function evaluated in step 6, etc., convergence is determined using a predetermined iterative calculation algorithm, and the decision variables 64, which are variables used in the iterative calculation, are updated. The predetermined algorithm can be an iterative method such as Newton's method, sequential quadratic programming, or the Levenberg-Marquardt method, and the designer can select an algorithm based on factors such as computational load and convergence.

In step 9, an estimated stiffness is determined based on the decision variable 64 for which the optimal solution was finally obtained through the iterative calculations of steps 4 to 8, and the stiffness information used in the control calculations for braking force control is updated.

Regarding the determination of the data range in step 2, the stiffness estimation shown in Figure 15 may be performed by accumulating data up until the brake is released as electric brake operation history and performing stiffness estimation after the brake is released. In this case, stiffness estimation can be performed when there is more data for stiffness estimation. Alternatively, stiffness estimation may be performed during braking when there is a sufficient amount of data to perform stiffness estimation while the electric brake is operating. In this case, stiffness estimation can be performed more quickly. Alternatively, stiffness estimation may be performed when there is a sufficient amount of data during braking, and then stiffness estimation may be performed again if there is more estimated data after the brake is released.

Figure 16 shows an example of a flow for storing estimated angles and estimated braking forces in the operation history storage unit 44 shown in Figures 7, 8, etc.

Step 10 determines whether the brakes are operating. This can be determined, for example, by determining whether an estimated braking force greater than a predetermined value is being generated.

In step 11, if it is determined in step 10 that braking is in progress, the current estimated angle and estimated braking force are stored.

Step 12, if it is determined in step 10 that braking is not occurring, determines whether the stored data is sufficient to perform stiffness estimation.

In step 13, if it is determined that the data stored in step 12 is insufficient for stiffness estimation, the stored data is reset. As a result, the stored data is reset, and the flow of steps 10 → 12 → 13 is executed until the next time the brakes are activated and braking force is generated.

In step 14, stiffness estimation is performed when it is determined that the data stored in step 12 is sufficient for stiffness estimation, and it is determined whether this stiffness estimation is complete. If stiffness estimation is complete, the data reset in step 13 is executed, and if it is not complete, no processing is performed until it is complete. In addition to the completion of all stiffness estimation processing, step 14 may also determine completion when, for example, the acquisition of stored data in the stiffness estimation processing is complete and there is no need to refer to the stored data until stiffness estimation is complete.

In step 13, for example, one cycle can be defined as the time from when braking force is generated to when the brake is released, and only stored data from before a certain cycle can be reset. Alternatively, only stored data for which a certain amount of time has elapsed can be reset. Alternatively, these processes can be used in combination as appropriate. Introducing these processes allows for the utilization of more acquired data, for example, when braking is performed repeatedly over a short period of time, when the stiffness of the electric brake device 1 is relatively unlikely to change.

Figure 17 shows an example of setting different limit ranges for the decision variable 64 during stiffness estimation, based on the range of change in the estimated braking force in the stored data used for stiffness estimation. Generally, if the range of change in the estimated braking force in the data used for stiffness estimation is small, it is necessary to estimate the overall stiffness from a partial braking operation range within the entire range. This may increase the risk of relatively large estimation errors occurring, particularly under conditions outside the data range. Therefore, as shown in Figure 17, it may be preferable to set limits on the decision variable 64 depending on the range of data used for stiffness estimation.

Step 15 derives the range of change in estimated braking force from the stored data used for stiffness estimation.

Step 16 sets a limit range for the decision variable 64 in stiffness estimation based on the range of change in the estimated braking force derived in step 15. The limit range for the decision variable 64 refers to the range of variation that can be achieved in the iterative calculations during stiffness estimation from the decision variable 64 that derived the estimated stiffness before stiffness estimation began. For example, if the reference address in the reference stiffness table 72 is the decision variable 64, the limit range is the range within which the address can vary from the initial address. Also, if the coupling coefficient is the decision variable 64, the limit range is the range within which it can vary from the initial coupling coefficient.

Figure 17 shows an image in which the limit range changes linearly within a specified braking force change range, and becomes constant at a certain upper limit value once the range exceeds the specified value. However, such an upper limit does not have to be set, and the limit range may change in a curved line relative to the braking change range.

In step 17, stiffness estimation shown in Figure 15 is performed. At this time, the decision variables 64 are iteratively calculated within the limit range determined in step 16 to search for the optimal solution.

Figure 18 shows an example in which the degree of brake use is derived from the sliding distance between the friction material 40 and the brake rotor 42 during braking and the braking force generated at that time, and a different limit range is set for the decision variable 64 during stiffness estimation based on the degree of brake use. In general, the characteristics of the friction material 40 are more likely to change the longer the braking distance with a larger braking force. For this reason, it may be preferable to vary the limit range of the decision variable 64 based on the sliding distance between the friction material 40 and the brake rotor 42 and the braking force.

Step 18 acquires the angular velocity of the brake rotor 42. For example, in the case of an electric brake device 1 for a vehicle, the angular velocity can be calculated from a wheel speed sensor such as an ABS sensor.

Step 19 calculates the sliding distance between the friction material 40 and the brake rotor 42 (hereinafter referred to as the "friction material sliding distance"). The friction material sliding distance can be derived from the angular velocity of the brake rotor 42 (brake rotor angular velocity) and time. Alternatively, for example, when data sampling is performed at regular intervals, the sliding distance per unit data sample is proportional to the angular velocity, so the angular velocity of the brake rotor 42 can be used directly as the equivalent sliding distance.

In step 20, the degree of brake use is derived from the friction material sliding distance and braking force based on a predetermined function g. This predetermined function g may be the product of the braking force and the friction material sliding distance. Alternatively, it is possible to determine in advance through experiments, etc., the extent to which a certain degree of friction material sliding occurs in the characteristics and shape of the friction material 40 when a certain braking force is used, and then set the predetermined brake use degree derivation function g based on the results of the experiment.

Step 21 sets a limit range for the decision variable 64 in stiffness estimation based on the degree of brake use derived in step 20.

Figure 19 shows an example of an electric brake device for an automobile where, when the conditions for performing stiffness estimation are not met for a long period of time, the electric brake device is operated under conditions that allow automatic stiffness estimation to be performed, thereby performing stiffness estimation. In an automobile, operation of the electric brake device 1 is basically left to the driver. For this reason, depending on the driving conditions, it is expected that the electric brake device 1 will operate without generating sufficient braking force to perform stiffness estimation, such as when light braking is repeated. For this reason, it may be necessary to perform the processing shown in Figure 19.

Step 22 determines whether stiffness estimation is being performed. If stiffness estimation is not being performed, the non-execution time is accumulated (step 23). If stiffness estimation is being performed, the accumulated time is reset (step 24).

If it is determined in step 22 that stiffness estimation is not being performed, step 25 determines whether the cumulative time in that state (cumulative time of non-execution) is greater than a predetermined value.

Step 26 acquires the vehicle's driving state when it is determined in step 25 that the cumulative non-execution time is greater than a predetermined value. Such vehicle driving state may be, for example, the vehicle speed of a vehicle equipped with electric brakes, or information from various sensors related to vehicle driving, such as an acceleration sensor, GPS, ABS sensors on multiple wheels, etc., mounted on the vehicle.

Step 27 determines whether the vehicle is stopped based on the vehicle driving status obtained in step 26.

In step 28, if it is determined in step 27 that the vehicle is stopped, the brake force target value is set to Fro regardless of the driver's operation. Here, "Fro" is a brake force target value that is at least greater than the brake force target value instructed by the driver while the vehicle is stopped, and is sufficient for performing stiffness estimation.

In step 22 of Figure 19, it may be determined whether stiffness estimation is being performed based on the electric brake operation history in which a predetermined brake force change has occurred. Even if stiffness estimation is being performed, if the brake force change in the electric brake operation history used for stiffness estimation is small, the estimation is based on limited information within the entire brake operation range. This increases the risk of relatively large estimation errors outside the operating range of the electric brake operation history. Therefore, it may be preferable to implement such processing.

Figure 20 shows an example of the operation of an electric brake device. Figure 20(a) shows an example in which the electric brake device 1 of this embodiment is applied. Figure 20(b) shows an example of a conventional electric brake device that does not have the configuration of this embodiment. In the example of this embodiment shown in Figure 20(a), the braking force remains constant even when the stiffness changes, and the accuracy of brake control is maintained. In contrast, in the conventional example shown in Figure 20(b), as a result of the change in stiffness of the electric brake device 1, the braking force also changes, and brake controllability deteriorates.

The present invention is not limited to the above-described embodiments, and various additions, modifications, and deletions are possible without departing from the spirit of the present invention. Therefore, such additions, modifications, and deletions are also included within the scope of the present invention.

1 Electric brake device 2 Control device 4 Direct acting mechanism (friction material operating means)
REFERENCE SIGNS LIST 10 Electric motor 12 Angle sensor 14 Load sensor 16 Reducer 20 Motion state estimator 22 Braking force estimator 24 Stiffness estimator 32 Angle estimator 34 Angular velocity estimator 36 Braking force control unit 38 Stiffness memory unit 40 Friction material 42 Brake rotor 46 Stiffness estimation calculation unit 50 Transmission mechanism 62 Reference stiffness 64 Decision variable

Claims (8)

An electric brake device comprising: a brake rotor; friction material that contacts the brake rotor to generate a braking force; friction material operation means that controls the contact state between the friction material and the brake rotor; an electric motor that powers the friction material operation means; and a control device that drives the electric motor to control the braking force caused by the contact between the friction material and the brake rotor,
The control device
an angle estimator that estimates a physical quantity corresponding to a rotation angle of the electric motor;
a braking force estimator that estimates a braking force;
a stiffness estimator that stores the estimated angle estimated by the angle estimator and the estimated braking force estimated by the braking force estimator as an electric brake operation history, and that estimates stiffness, which is a correlation that associates the rotation amount of the electric motor with the braking force, based on the stored estimated angle and estimated braking force;
a braking force control unit that derives a motor drive amount for causing the estimated braking force to follow a target value of the braking force using the estimated stiffness estimated by the stiffness estimator,
the stiffness estimator has a stiffness estimation calculation unit that calculates an estimated stiffness using a plurality of reference stiffnesses stored in advance and a decision variable that determines one of the reference stiffnesses or an intermediate value thereof,
the stiffness estimation calculation unit compares a calculated value derived from one of the stored estimated angle and the stored estimated braking force using the estimated stiffness with the other stored data to derive an error, and estimates stiffness by performing a convergence calculation that adjusts the decision variable so that the error becomes smaller to at least a predetermined allowable amount ;
the plurality of reference stiffnesses stored in the stiffness estimator are at least two or more patterns of different reference stiffnesses,
An electric brake device in which the decision variable is a value multiplied as a combined ratio of a plurality of the reference stiffnesses . 2. The electric brake device according to claim 1, wherein the plurality of reference stiffnesses stored in the stiffness estimator are a data table having reference stiffnesses at a plurality of addresses, respectively;
An electric brake device in which the decision variable is an address indicating a reference destination of the data table. An electric brake device comprising: a brake rotor; friction material that contacts the brake rotor to generate a braking force; friction material operation means that controls the contact state between the friction material and the brake rotor; an electric motor that powers the friction material operation means; and a control device that drives the electric motor to control the braking force caused by the contact between the friction material and the brake rotor,
The control device
an angle estimator that estimates a physical quantity corresponding to a rotation angle of the electric motor;
a braking force estimator that estimates a braking force;
a stiffness estimator that stores the estimated angle estimated by the angle estimator and the estimated braking force estimated by the braking force estimator as an electric brake operation history, and that estimates stiffness, which is a correlation that associates the rotation amount of the electric motor with the braking force, based on the stored estimated angle and estimated braking force;
a braking force control unit that derives a motor drive amount for causing the estimated braking force to follow a target value of the braking force using the estimated stiffness estimated by the stiffness estimator,
the stiffness estimator has a stiffness estimation calculation unit that calculates an estimated stiffness using a plurality of reference stiffnesses stored in advance and a decision variable that determines one of the reference stiffnesses or an intermediate value thereof,
the stiffness estimation calculation unit compares a calculated value derived from one of the stored estimated angle and the stored estimated braking force using the estimated stiffness with the other stored data to derive an error, and estimates stiffness by performing a convergence calculation that adjusts the decision variable so that the error becomes smaller to at least a predetermined allowable amount;
an electric brake device, wherein the decision variables in the stiffness estimation calculation unit include a first decision variable that mainly changes nonlinearity of stiffness and a second decision variable that mainly changes overall stiffness; An electric brake device comprising: a brake rotor; friction material that contacts the brake rotor to generate a braking force; friction material operation means that controls the contact state between the friction material and the brake rotor; an electric motor that powers the friction material operation means; and a control device that drives the electric motor to control the braking force caused by the contact between the friction material and the brake rotor,
The control device
an angle estimator that estimates a physical quantity corresponding to a rotation angle of the electric motor;
a braking force estimator that estimates a braking force;
a stiffness estimator that stores the estimated angle estimated by the angle estimator and the estimated braking force estimated by the braking force estimator as an electric brake operation history, and that estimates stiffness, which is a correlation that associates the rotation amount of the electric motor with the braking force, based on the stored estimated angle and estimated braking force;
a braking force control unit that derives a motor drive amount for causing the estimated braking force to follow a target value of the braking force using the estimated stiffness estimated by the stiffness estimator,
the stiffness estimator has a stiffness estimation calculation unit that calculates an estimated stiffness using a plurality of reference stiffnesses stored in advance and a decision variable that determines one of the reference stiffnesses or an intermediate value thereof,
the stiffness estimation calculation unit compares a calculated value derived from one of the stored estimated angle and the stored estimated braking force using the estimated stiffness with the other stored data to derive an error, and estimates stiffness by performing a convergence calculation that adjusts the decision variable so that the error becomes smaller to at least a predetermined allowable amount;
the friction material operating means is a linear motion mechanism equipped with a speed change mechanism in which the rotational motion of the electric motor is converted into linear motion, and an equivalent lead, which is the correlation between the amount of rotation and the amount of linear motion, changes at a predetermined braking force;
the plurality of reference stiffnesses stored in the stiffness estimator are reference stiffnesses including changes in the equivalent lead;
An electric brake device, wherein the decision variables include a decision variable that mainly changes a braking force condition under which the change in the equivalent lead occurs. 5. The electric brake device according to claim 4 , wherein the friction material operating means comprises a planetary reduction structure having a rotation input member and planetary rolling elements arranged coaxially with a rotation axis of the rotation input member and equidistantly spaced apart in the circumferential direction, and which produces a reduction effect depending on the ratio between the revolution speed of the rotation input member and the revolution speed of the planetary rolling elements,
an elastic member that applies a fastening force that rotates the rotation input member and the planetary rolling element integrally;
The friction material is coupled to the brake rotor through a reaction force, and the elastic member is released from the coupling force, resulting in a planetary deceleration effect.
the plurality of reference stiffnesses stored in the stiffness estimator are reference stiffnesses including deformation amounts of the elastic member,
an electric brake device, wherein the decision variables in the stiffness estimation calculation unit include a decision variable that changes a braking force condition at which deformation of the elastic member is completed; An electric brake device comprising: a brake rotor; friction material that contacts the brake rotor to generate a braking force; friction material operation means that controls the contact state between the friction material and the brake rotor; an electric motor that powers the friction material operation means; and a control device that drives the electric motor to control the braking force caused by the contact between the friction material and the brake rotor,
The control device
an angle estimator that estimates a physical quantity corresponding to a rotation angle of the electric motor;
a braking force estimator that estimates a braking force;
a stiffness estimator that stores the estimated angle estimated by the angle estimator and the estimated braking force estimated by the braking force estimator as an electric brake operation history, and that estimates stiffness, which is a correlation that associates the rotation amount of the electric motor with the braking force, based on the stored estimated angle and estimated braking force;
a braking force control unit that derives a motor drive amount for causing the estimated braking force to follow a target value of the braking force using the estimated stiffness estimated by the stiffness estimator,
the stiffness estimator has a stiffness estimation calculation unit that calculates an estimated stiffness using a plurality of reference stiffnesses stored in advance and a decision variable that determines one of the reference stiffnesses or an intermediate value thereof,
the stiffness estimation calculation unit compares a calculated value derived from one of the stored estimated angle and the stored estimated braking force using the estimated stiffness with the other stored data to derive an error, and estimates stiffness by performing a convergence calculation that adjusts the decision variable so that the error becomes smaller to at least a predetermined allowable amount;
the stiffness estimation calculation unit determines a brake operation amount based on at least one of an amount of change in an estimated brake force and an amount of change in an estimated angle in the electric brake operation history, and when the brake operation amount becomes small, limits a change in the decision variable in a convergence calculation when estimating the stiffness. An electric brake device comprising: a brake rotor; friction material that contacts the brake rotor to generate a braking force; friction material operation means that controls the contact state between the friction material and the brake rotor; an electric motor that powers the friction material operation means; and a control device that drives the electric motor to control the braking force caused by the contact between the friction material and the brake rotor,
The control device
an angle estimator that estimates a physical quantity corresponding to a rotation angle of the electric motor;
a braking force estimator that estimates a braking force;
a stiffness estimator that stores the estimated angle estimated by the angle estimator and the estimated braking force estimated by the braking force estimator as an electric brake operation history, and that estimates stiffness, which is a correlation that associates the rotation amount of the electric motor with the braking force, based on the stored estimated angle and estimated braking force;
a braking force control unit that derives a motor drive amount for causing the estimated braking force to follow a target value of the braking force using the estimated stiffness estimated by the stiffness estimator,
the stiffness estimator has a stiffness estimation calculation unit that calculates an estimated stiffness using a plurality of reference stiffnesses stored in advance and a decision variable that determines one of the reference stiffnesses or an intermediate value thereof,
the stiffness estimation calculation unit compares a calculated value derived from one of the stored estimated angle and the stored estimated braking force using the estimated stiffness with the other stored data to derive an error, and estimates stiffness by performing a convergence calculation that adjusts the decision variable so that the error becomes smaller to at least a predetermined allowable amount;
further comprising an angular velocity estimation unit that estimates the angular velocity of the brake rotor,
the stiffness estimation calculation unit determines a degree of brake use based on at least one of the time period during which the estimated angle and estimated brake force of the electric brake operation history are acquired, the angular velocity of the brake rotor, and the estimated brake force, and is configured to limit changes in the decision variable in a convergence calculation when estimating the stiffness when the degree of brake use becomes small. An electric brake device comprising: a brake rotor; friction material that contacts the brake rotor to generate a braking force; friction material operation means that controls the contact state between the friction material and the brake rotor; an electric motor that powers the friction material operation means; and a control device that drives the electric motor to control the braking force caused by the contact between the friction material and the brake rotor,
The control device
an angle estimator that estimates a physical quantity corresponding to a rotation angle of the electric motor;
a braking force estimator that estimates a braking force;
a stiffness estimator that stores the estimated angle estimated by the angle estimator and the estimated braking force estimated by the braking force estimator as an electric brake operation history, and that estimates stiffness, which is a correlation that associates the rotation amount of the electric motor with the braking force, based on the stored estimated angle and estimated braking force;
a braking force control unit that derives a motor drive amount for causing the estimated braking force to follow a target value of the braking force using the estimated stiffness estimated by the stiffness estimator,
the stiffness estimator has a stiffness estimation calculation unit that calculates an estimated stiffness using a plurality of reference stiffnesses stored in advance and a decision variable that determines one of the reference stiffnesses or an intermediate value thereof,
the stiffness estimation calculation unit compares a calculated value derived from one of the stored estimated angle and the stored estimated braking force using the estimated stiffness with the other stored data to derive an error, and estimates stiffness by performing a convergence calculation that adjusts the decision variable so that the error becomes smaller to at least a predetermined allowable amount;
the control device has a driving state estimator that estimates a driving state of a vehicle equipped with the electric brake device,
The stiffness estimation calculation unit
a function of determining whether or not stiffness estimation has been performed based on an electric brake operation history in which a change in estimated brake force greater than a predetermined value has occurred, and measuring a non-execution time during which stiffness estimation has not been performed based on the determination;
an electric brake device having a function of generating a braking force equal to or greater than the predetermined braking force, regardless of operation by the driver of the vehicle, when the non-execution time has elapsed for a predetermined period of time or more and the driving state is estimated to be that the vehicle on which the electric brake device is mounted has been stopped for a predetermined period of time or more and the braking force of the vehicle is smaller than a predetermined value, and performing stiffness estimation using the estimated braking force and estimated angle at that time.