JP7202567B2 - Multi-body minimum mechanical parameter identification device, method and program - Google Patents

Description

The present invention relates to a multi-body minimum dynamics parameter identification device, method, and program.

An object has a unique inertial property consisting of a total of 10 parameters: a mass consisting of one parameter, a center of gravity consisting of three parameters, and an inertia tensor consisting of six parameters necessary for dynamic behavior analysis and evaluation. is defined. Note that the inertia tensor consists of the moment of inertia (diagonal term of inertia) and the product of inertia. For example, the position of the center of gravity is an important parameter that leads to rollover prevention in vehicles such as automobiles and heavy machinery. In general, in vehicles such as automobiles, not only the inertia diagonal term but also the inertia product greatly affects the evaluation of maneuverability, comfort, motion performance, and the like.

As shown in FIG. 1, a conventional unit model inertia characteristic identification unit based on free vibration measurement includes a platform 101 on which a unit model 100 is mounted, and a pseudo-supporting platform 101 that allows free vibration (pseudo peripheral It consists of a support member (spring/wire) 102 (called support under free boundary conditions) and a sensor 103 (for example, an acceleration sensor) for detecting vibration of the platform 101 (see: Patent Documents 1 and 2). In this case, the platform 101 and the single model 100 are regarded as an integral rigid body. to obtain the frequency response at . By accurately identifying the inertia characteristics of the platform 101 in advance and subtracting the platform 101 from the integral rigid body characteristics of the platform 101 and the single model 100, the inertia characteristics of the single model 100 can be obtained. In this way, the effects of the support member 102 such as springs, wires, etc. can be modeled and removed to enable high-precision identification. Since it can be identified only by grasping the value of the vibration frequency, there is no need to calibrate the sensitivity of the sensor, and the measurement error is small. An advanced version of this single model inertia property identification unit is used as the multibody model overall inertia property measurement unit according to the present invention.

On the other hand, in robotics, ergonomics, etc., geometric parameters such as the length of the constituent objects (links) and each constituent object (link ) inertial characteristics are required. In this case, if there is CAD data for each constituent object, each inertial characteristic can be obtained by calculation using a CAD model. However, in a mechanical mechanism (for example, a robot) in an assembled state or a body in a state that cannot be separated, the identification of each inertial characteristic can only be done by experimental measurement. In doing so, the geometric parameters are easily identifiable from static measurements, but since the inertial properties of each constituent body (link) are redundant to the multi-body dynamic model, the motion of the constituent bodies (links) and cannot be identified from joint torque or external force data. Therefore, the minimum dynamic parameters are defined as the minimum dynamic parameters required to express the dynamic model of the multi-body system.

A first conventional method for identifying the minimum dynamic parameters of a multi-body system is to provide a torque sensor at each joint in a multi-body system in which a plurality of links are connected by joints, and measure the motion with respect to the joint torque. Identify the minimum dynamic parameters of the inertial properties of (see: Non-Patent Document 1).

The second conventional multi-body minimum mechanical parameter identification method identifies the minimum mechanical parameter of each constituent object (link) by measuring the motion and the ground reaction force against the motion (see Non-Patent Document 2).

JP 2012-18092 A Japanese Patent Publication No. 2014-515492

H. Kawasaki: Parameter Identification of Robotic Arms, Measurement and Control, Vol.28, No.4, pp. 344-350, 1989. K. Ayusawa, G. Venture, Y. Nakamura : Minimal-set of inertial parameters identification of legged robots using base-link dynamics, Journal of the Robotics Society of Japan, Vol. 27 No. 9, 2009. R. Klopper : A measurement system for rigid body properties enabled by gravity-dependent suspension modeling, PhD.Thesis, 2009. R. Klopper, M. Okuma : Experimental identification of rigid body inertia properties using low-frequency unbalance excitation, Proc. 50th ASME Structural dynamics conference, 2007. R. Klopper, H. Akita, M. Okuma, S. Terada : An experimental identification method for rigid body properties enabled by gravity-dependent suspension modeling, Proceedings of joint inter-national conference on multibody system dynamics, 2010.

However, the above-described first conventional multi-body minimum dynamics parameter identification method requires highly accurate measurement of the force of torque, and therefore has a large limitation in its application. For example, in the analysis of human movement in sports science and engineering, and in the analysis of humanoid robots, it is difficult to place torque sensors at all joints. is difficult, there is a problem that the restrictions of application are large.

Also, in the above-described second conventional multi-body minimum dynamics parameter identification method, since it is necessary to measure the force of the floor reaction force, application is greatly restricted. In other words, it is difficult to measure both the drag force and the frictional force as floor reaction forces with high accuracy, and it is difficult to select the optimum motion in measuring three-dimensional motion using motion capture. Since it is difficult to generalize to the identification of three-dimensional inertial characteristics, there is a problem that application is greatly restricted.

In order to solve the above-described problems, a multi-body minimum dynamics parameter identification device according to the present invention mounts a multi-body model in which a plurality of objects are connected under a pseudo peripheral free boundary condition, and mounts the multi-body model A global inertial property measurement unit that detects the natural angular frequency of free vibration of the system, and a multi-body model that is connected to the global inertial property measurement unit, is mounted under pseudo-peripheral free boundary conditions, and measures the internal force between links. A relative motion measurement unit for measuring the vibration or motion by making it move relative to each other, a global inertial property measurement unit and a relative motion measurement unit, and the global inertial property of the multi-body model in multiple poses and its vibration or motion and a control unit for establishing a system of equations based on to identify the minimum mechanical parameters of each object. It should be noted that the overall inertial characteristic measurement unit and the relative motion measurement unit can be the same as hardware, or can be prepared as separate units.

In addition, the multi-body minimum dynamic parameter identification method according to the present invention is a multi-body model in which a plurality of objects are connected is mounted on a global inertial property measurement unit under pseudo peripheral free boundary conditions, and free vibration in multiple postures is performed. and the relative motion for measuring the vibration or motion by mounting the multi-body model under pseudo peripheral free boundary conditions and making it move relative to the internal force between links. a measurement step; and a minimum mechanical parameter identification step for identifying the minimum mechanical parameter of each object by coupling equations based on the global inertial properties of the multi-body model at multiple poses and its vibration or motion. .

Furthermore, the multi-body minimum dynamic parameter identification program according to the present invention detects the natural angular frequency of a multi-body model in which a plurality of objects are connected under pseudo peripheral free boundary conditions, Overall inertia characteristics measurement procedure for measuring the overall inertia characteristics of the system, and relative motion measurement for measuring the vibration or motion by mounting the multi-body model under pseudo free boundary conditions and making it move relative to the internal force between links. and a minimum mechanical parameter identification procedure for identifying the minimum mechanical parameters of each body by coupling equations based on the global inertial properties and their vibration or motion.

According to the present invention, it is possible to identify the minimum mechanical parameters of a multi-body system without force measurement.

1A and 1B are a front view and a perspective view, respectively, showing an example of a conventional single model inertial characteristic identification unit. FIG. 1 is a block diagram showing an embodiment of a multi-body minimum dynamics parameter fixing device according to the present invention; FIG. FIG. 3 is a diagram showing an example of the multi-body model of FIG. 2; FIG. 3 shows an example of a commercially available product of the total inertial characteristic measurement unit of FIG. 2, (A) is a front view, and (B) is a perspective view. An example of the relative motion measurement unit of FIG. 2 is shown, (A) is a front view, (B) is a perspective view. 3 is a flow chart for explaining the operation of the control unit of FIG. 2; FIG. 7 is a perspective view showing a multi-body model for explaining the posture changing step of FIG. 6; FIG. 7 is a perspective view showing a plurality of postures, eg, three postures, of the multibody model in the posture changing step of FIG. 6; FIG. 7 is a detailed flowchart of a global inertial property measurement step 602 of FIG. 6; FIG. 7A is a perspective view showing a measurement state, and FIG. 7B is a diagrammatic representation of measurement data. FIG. FIG. 9 is a table showing overall inertia characteristic measurement results (average value of seven measurements) for each posture in FIG. 8; FIG. 12A and 12B are diagrams showing the identification result of the minimum dynamic parameter regarding the position of the center of gravity based on the measurement result of FIG. 11 and its accuracy; FIG. 11B is a diagram showing the identification result and accuracy of the minimum dynamic parameter regarding the inertia tensor of the base link based on the measurement data of FIG. 10B and the measurement result of FIG. 11; 11. It is a figure for demonstrating the precision evaluation of the identification result of the minimum dynamics parameter regarding the inertia tensor of the link 1 based on the measurement data of FIG.10(B), and the measurement result of FIG. FIG. 5 is a diagram for explaining another embodiment of the present invention; It is a figure explaining an example of a field to which the present invention is applied.

Bold and italic numbers and letters of vectors and matrices in Equation 1 below are normal numbers and letters in the specification.

FIG. 2 is a block diagram showing a multi-body minimum dynamics parameter identification device according to the present invention.

In FIG. 2, an overall inertia property measurement unit 20 for measuring the overall inertia property in multiple postures of the multi-body model 10 and a relative motion measurement unit 30 for measuring the relative motion between each link of the multi-body model 10. is provided, and a control unit 40 (for example, a microcomputer) for controlling the overall inertial property measurement unit 20 and the relative motion measurement unit 30 is provided. As for the hardware configuration, the overall inertial characteristic measurement unit 20 and the relative motion measurement unit 30 can be unified in design.

FIG. 3 is a diagram showing an example of the multibody model 10 of FIG.

FIG. 3 shows a tree structure of N links j, where link 0 is the base link, but any link can be the base link. In the case of a structural model of a human trunk and limbs, the trunk link is the base link.

FIG. 4 shows an example of the overall inertial characteristic measurement unit 20 of FIG. 2, (A) is a front view, and (B) is a perspective view. The overall inertia characteristic measurement unit 20 shown in FIG. 4 is a development of the conventional single model inertia characteristic identification unit shown in FIG.

In FIG. 4, a platform 21 for mounting the multi-body model 10, a support member (spring) 22 supporting the platform 21 so that it can freely vibrate, that is, under a pseudo-peripheral free boundary condition, and the vibration of the platform 21 are detected. A strain sensor 23 is provided for the purpose. The strain sensor 23 is connected to the control unit 40 . In this case as well, the platform 21 and the multibody model 10 are regarded as an integral rigid body, and the integral rigid body is vibrated by a vibration exciter, impulse hammer, or the like to obtain a frequency response in a low frequency range where no elastic mode occurs. . By accurately identifying the inertia characteristics of the platform 21 in advance and subtracting the inertia characteristics of the platform 21 from the integral inertia characteristics of the platform 21 and the multibody model 10, the inertia characteristics of the multibody model 10 can be obtained. In this way, the effects of the support members 22 such as springs, wires, etc. can be modeled and removed to enable high-precision identification. Since it is possible to identify only the value of the vibration frequency, the error in the identification result of the inertial characteristics of the multibody model 10 is also small.

FIG. 5 shows an example of the relative motion measurement unit 30 of FIG. 2, (A) is a front view, and (B) is a perspective view.

In FIG. 5, a platform 31 for mounting the multi-body model 10, a support member (wire) 32 supporting the platform 31 under a pseudo peripheral free boundary condition with respect to horizontal minute displacement motion, and relative motion of each link are detected. A motion capture 33 is provided for Motion capture 33 is connected to control unit 40 .

ここで、ｉ＝Ｏ、Ｃとおいて、
Ｈ_i,j（ｊ＝１、２）は慣性行列
ｑ_ｏはベースリンクの一般化座標（６自由度）
ｑ_Ｃは関節角度ベクトル
ｂ_ｉコリオリ力、遠心力、重力を表すベクトル項
τは関節トルクを表すベクトル
Ｎ_Ｃは多体系に働く外力の総数
Ｆ_ｋは多体系に働くｋ番目の外力ベクトル
Ｊ_ｉｋはｋ番目の外力を一般化力へ変換する行列 The equation of motion of the multibody model 10 in FIG.

Here, with i = O, C,
H _i,j (j=1, 2) is the inertia matrix q _o is the generalized coordinates of the base link (6 degrees of freedom)
q _C is the joint angle vector b _i representing the Coriolis force, centrifugal force, and gravity Vector term τ is the vector representing the joint torque NC is the total number of external forces acting on the multi _- body system F _k is the k-th external force vector J _ik acting on the multi-body system is a matrix that transforms the k-th external force into a generalized force

φは数３式に示すＮ個のリンクｊの力学パラメータを並べたベクトルである。

φ_jは数４式に示すリンクｊの１０個の力学パラメータによって構成されるベクトルである。

ここで、ｍ_ｊはリンクｊの質量
Ｉ_ｊ，ｘｘ，Ｉ_ｊ，ｙｙ，Ｉ_ｊ，ｚｚ，Ｉ_ｊ，ｘｙ，Ｉ_ｊ，ｘｙ，Ｉ_ｊ，ｙｚはリンクｊの座標原点回りの慣性テンソルＩ_ｊの各成分
ｇ_ｊ，ｘ，ｇ_ｊ，ｙ，ｇ_ｊ，ｚはリンクｊの座標原点からの重心位置ベクトルｇ_ｊの各成分
Ｙｉはφ_ｊにかかるリグレッサ行列と呼ばれ、ｑ_ｊ、ｑ_ｉの１階時間微分及びｑ_ｉの２階時間微分に基づいて構成される関数行列 The equation of motion of Equation 1 can be transformed as shown in Equation 2, and has a linear relationship with the dynamic parameters of mass, center of gravity, and inertia tensor.

φ is a vector in which the dynamic parameters of N links j shown in Equation 3 are arranged.

φ _j is a vector composed of 10 dynamic parameters of link j shown in Equation (4).

where m _j is the mass of link j I _{j, xx} , I _{j, yy} , I _{j, zz} , I _{j, xy} , I _{j, xy} , and I _{j, yz} is the inertia tensor I around the coordinate origin of link j Each component g _{j ,x} , g _j,y , g _j,z of _j is the center of gravity position vector g _j from the coordinate origin of link _j . A function matrix constructed based on the first time derivative of _i and the second time derivative of q _i

ここで、Ｙ_ｉＢは最小力学パラメータベクトルφ_Ｂのリグレッサ行列 Not all dynamic parameters can be identified from Equation 2. This is because the inertial characteristics of each link j are redundant with respect to the multibody model 11 . Therefore, the expression 2 can be transformed into the expression 5 by introducing the vector φ _B consisting of the minimum mechanical parameters.

where Y _iB is the regressor matrix of the minimum dynamics parameter vector φ _B

In the present invention, using the equation of motion that does not include torque τ in the upper part of Equation 5, geometric parameters such as the link length and the position of the pair, the motion of each link j, the spring of the measurement unit, and the attachment point of the wire to the measurement table Substitute data obtained by measuring position and motion to create an equation for obtaining the minimum mechanical parameters. Therefore, there is no need to measure the joint torque of the multibody model. In addition, the forces generated in springs and wires are precisely modeled in advance with the mechanical parameters (spring constant and length) of springs and wires, so they can be calculated from position measurements, and external forces such as floor reaction force can be measured. do not need. The solution of this equation is combined with the result of identifying the inertial characteristics of the entire multi-body model 10 in the basic attitude state when relative motion is made between the links j.

Next, the operation of control unit 40 of FIG. 2 will be described with reference to FIG.

First, in step 601, the multi-body model 10 is mounted on the platform 21 of the global inertial property measurement unit 20, and the attitude of the multi-body model 10 is changed. Posture is fixed after change. Assuming that the multi-system model 10 is a two-link system as shown in FIG. 7, the postures shown in FIGS. 8A, 8B, and 8C are assumed.

Next, in step 602, the overall inertial characteristics of the multibody model 10 in a fixed posture are measured.

Next, at step 603, it is determined whether or not the measurement of the overall inertia characteristics for the attitude change is finished. That is, steps 601 and 602 are repeated until all the measurements in the postures shown in FIGS. 8A, 8B, and 8C are completed.

In step 604, based on the overall inertia characteristics in multiple postures obtained in step 602, the minimum dynamics parameter regarding the mass of each link, the minimum dynamics parameter regarding the center of gravity of each link, and the product of inertia of the inertia tensor of each link are calculated. Identify the minimum mechanical parameters.

Next, in step 605, the multibody model 10 is mounted on the platform 31 of the relative motion measurement unit 30 of FIG. 5, and the relative motion measurement between the links j by internal force is performed.

Next, in step 606, an equation (the upper part of equation 5 ) to identify the minimum dynamic parameters for the diagonal components of the inertia tensor of each link j.

Then, at step 607, the routine of FIG. 6 ends.

FIG. 9 is a detailed routine of the overall inertial characteristic measurement step 602 in FIG.

First, in step 901, the platform 21 and the multibody model 10 are vibrated to obtain a plurality of natural angular frequencies from the free vibration response. In addition, since the number of natural angular frequencies that can be observed may differ depending on how the free vibration is applied, it is important to vibrate so that six or as many as possible can be observed during excitation. .

Next, at step 902, the initial value of the unknown inertia characteristic is determined.

Next, in step 903, the mass matrix M of the platform 21 and the multi-body model 10 as a unit is created from the inertia characteristic values. Note that the mass matrix is a 6-row by 6-column matrix composed of the mass, the center of gravity, and the inertia tensor.

Next, at step 904, a stiffness matrix K is created. A stiffness matrix K can be constructed from the spring constants and the end positions of the springs attached between the fixed support position and the platform 21 . The stiffness matrix K is a 6-row by 6-column matrix representing the force applied to the movable part (the multibody model 10 and the platform 21) when the movable part (the multibody model 10 and the platform 21) is slightly displaced in translation and rotation from the static balanced state.

In step 905, the static equilibrium state of the platform 21 and the multi-system model 10 as a unit is calculated by numerical analysis. The static balance state can be solved based on the static balance equation between the stiffness matrix K obtained in step 904 and the mass matrix M and gravity of the single unit determined in step 903 .

Next, in step 906, from the mass matrix M obtained in step 903 and the stiffness matrix K obtained in step 905, the theoretical eigenvalue analysis of the minute free vibration at the statically balanced position obtained in step 905 is performed. By constructing and solving the equation, the square root of its eigenvalue is calculated as the six theoretical natural angular frequencies ω. in short,
det(K ^- ω2M)=0
Calculate the theoretical natural angular frequency ω from However, I is the identity matrix.

Next, in step 907, the square norm of the difference between the experimental natural angular frequency obtained in step 901 and the theoretical natural angular frequency obtained in step 906 is calculated.

Next, in step 908, it is determined whether the squared norm obtained in step 907 is smaller than the previous value of the iterative calculation from steps 903 to 908. If so, sensitivity analysis is performed in the direction of convergence, and based on this, the inertial characteristics are appropriately slightly changed and steps 903 to 907 are repeated. When it is determined that the square norm converges to a sufficiently small value, the overall inertia characteristic is identified from the mass matrix value at that time.

Then, at step 909, the routine of FIG. 9 ends.

ここで、下添字l（エル）はl（エル）回目の計測を示す
m_s、g_s、I_sは多体系全体の質量、ベースリンク座標原点からの重心位置ベクトルおよび慣性テンソル
p_jはベースリンク原点からリンクj座標原点へ向かうベクトル（j≧1）
R_jはベースリンク原点からリンクj座標原点への姿勢行列（j≧1）
[ω×] は数８式のような演算子

In step 905 of FIG. 9, the stiffness matrix K is derived by modeling the effects of springs and wires on minute rigid body motions from equilibrium. Influence factors of springs and wires can be considered to be the following three.
1. Spring restoring force (spring axial restoring force)
2. Geometric stiffness, such as pendulum motion, that tends to return to its original position due to gravity when it deviates from its equilibrium position.
3. Restoring moment (restoring force in the direction perpendicular to the axis of the spring) when the balance of the moment is lost from the equilibrium position
See Non-Patent Documents 3, 4, and 5 for detailed derivation of the mass matrix M and stiffness matrix K. In this way, by using springs and wires whose characteristic values have been accurately grasped in advance and precisely modeling their influence, high-accuracy identification has become possible. If this identification method is used to identify the overall inertia characteristics in a plurality of postures, in step 604 in FIG. can be done.

where the subscript l (ell) indicates the l (ell) measurement
m _s , g _s , and I _s are the mass of the whole many-body body, the center-of-gravity position vector from the base-link coordinate origin, and the inertia tensor
p _j is a vector from the base link origin to the link j coordinate origin (j ≥ 1)
R _j is the attitude matrix from the base link origin to the link j coordinate origin (j≧1)
[ω×] is an operator like Equation 8

From Equations 6 and 7, it is possible to obtain the relational expressions relating to the minimum mechanical parameter regarding the position of the center of gravity, the minimum mechanical parameter regarding the corresponding portion regarding the product of inertia of the inertia tensor, and the minimum mechanical parameter regarding the diagonal term component of the inertia tensor. For example, in a multi-system system in which the coordinate origin of the base link is set at the center of the spherical pair in the two-link system shown in FIG. Equation 10 is expressed by equation 10, and equations relating to each link and the entire inertia tensor are expressed by equation 11.

By solving the simultaneous equations of _Equation 10 _, m _j g _j can be obtained, and from the simultaneous equations of _Equation 11 and the properties of the rotation matrix, get _j,xx - I _j,yy , I _j,yy - I _j,zz , I _0,xx + I _1,xx , I _0,yy + I _1,yy , I _0,zz + I _1,zz be able to. Therefore, from a relatively simple action, such as the relative motion of a link in a plane in step 605 of FIG. are identifiable.

Next, the relative motion measurement step 605 between links by internal force in FIG. 6 will be described in detail.

In order to obtain the minimum dynamics parameter for the diagonal term component of the remaining inertia tensor, as shown in FIG. A movement is caused and the movement is measured. When performing measurement with the motion capture 33 (FIG. 5), the translation vector and rotation matrix of base link 0 and the time history change of the rotation matrix of each link are obtained from the position of each marker attached on each link in advance, and the translation vector And the velocity, acceleration, angular velocity, and angular acceleration vector are obtained from the time history change of the rotation matrix. By approximating the time history change of each component of the translation vector and rotation matrix to a polynomial that is the optimum approximation for each component using the least-squares method, and differentiating the polynomial, the velocity, acceleration, angular velocity, and angular acceleration vectors can be obtained. Obtain time history changes to improve the stability of numerical analysis and the accuracy of results. In this measurement method, the reaction force generated by the springs and wires of the measurement system installed in the relative motion measurement unit in order to create a pseudo peripheral free boundary condition on the base link 0 is an external force, so the stiffness matrix K is used. , the upper equation that does not include torque in Equation 5 can be expressed as Equation 12.

ただし、

ω_jは絶対座標系でのリンクjの角速度ベクトル
R₀は絶対座標原点からベースリンク座標原点への姿勢行列 Equation 12 is an equation for a total of 6 degrees of freedom, 3 degrees of freedom in translation and 3 degrees of freedom in rotation. The expression for translation consists of minimum mechanical parameters for total mass and center of gravity position. Therefore, the minimum mechanical parameters related to the overall mass and center of gravity position have already been identified from the overall inertial property measurement, so they are not required. By using one or more of the equations related to rotation in Equation 12, a part of the minimum dynamic parameters obtained in the overall inertia characteristic identification, and the relational expressions, the minimum dynamic parameters related to the inertia diagonal term component of the inertia tensor can be obtained. , all the minimum mechanical parameters can be identified. For example, in the case of FIG. 7, Equation 12 becomes Equation 13.

however,

ω _j is the angular velocity vector of link j in the absolute coordinate system
R ₀ is the attitude matrix from the absolute coordinate origin to the base link coordinate origin

In this way, m _j g _j , I _{j, xy} , I _{j, xz} , I _j,yz , I _j,xx − I _j,yy , I _j,yy − I _j,zz , I _0,xx + I _1,xx , I _0,yy + I _1,yy , I _0,zz + I By substituting _1,zz , we can obtain the minimum dynamics parameter for the diagonal elements of the inertia tensor.

FIG. 11 shows the overall inertia characteristic results and rotation matrix components obtained by averaging seven measurements in each of the postures A, B, and C in FIG. Based on the measurement data of FIG. 10(B) and the measurement results of FIG. 11, FIG. , and FIG. 14 shows the experimental identification results for the inertia tensor of link 1 .

As shown in FIG. 12, the maximum error component of the experimentally identified parameters m _j g _j composed of the mass and the position of the center of gravity is ±0.0028 kgm, which is ±0.0004 m per unit mass of the base link 0 . This is within the error range ±RMSE defined by the root mean square error (RMSE) when identifying 10,000 times in consideration of the random noise level theoretically considered to be included in the measurement data in the simulation.

As shown in FIGS. 13 and 14, the error in the experimental identification result of the inertia tensor is ±3.1%, which is defined by the root mean square error RMSE when identification is performed 10,000 times in the simulation as described above. within the error range ±RMSE.

6 and 9 are stored in the memory of the control unit 40 as a program.

Also, in the above-described embodiment, the multi-system system consisting of two links was taken as an example, but the present invention can also be applied to a multi-system system consisting of three or more links. For example, in the case of identifying the minimum dynamic parameters of an individual human being as a multi-body system, as shown in FIG. Perform relative motion measurement.

Furthermore, in the field of sports engineering, it has been impossible to identify the minimum mechanical parameters of an individual human until now. It is possible to calculate the burden (joint torque, etc.) on the joints of each player, and to construct an evaluation value for preventing injuries in advance. In addition, it is possible to propose an exercise that fits the person, such as what kind of movement the person with the minimum mechanical parameter should do.

Furthermore, in the automobile industry, automobile design is expected to suit each driver. For that purpose, it is pointed out that it is useful to conduct tests using a dummy doll shown in FIG. ing. Minimum mechanical parameters of a living human individual are required, but until now have been considered unidentifiable. By applying the present invention, an individual's minimum mechanical parameters can be identified, making it possible to develop corresponding dummies.

Furthermore, in automotive engineering, it would be ideal to be able to identify not only the inertia of the entire car, but also the inertia of the body and the inertia of the engine (or power-train) without disassembling the car. . As shown in FIG. 16(C), a mechanical structure in which a plurality of components are coupled by such elastic coupling is referred to as a "multi-body structural system". Identifiable without decomposition.

Also, the present invention can be applied to any modifications within the obvious scope of the above-described embodiments.

INDUSTRIAL APPLICABILITY As described above, the present invention can be used in fields such as sports engineering and automobile engineering.

10: Multibody model 0: Base links 1, 2, . : Control unit 100: Single model 101: Platform 102: Support member 103: Acceleration sensor

Claims (9)

A multi-body model in which a plurality of objects are connected is mounted under a pseudo peripheral free boundary condition, and the natural angular frequency of the free vibration of the multi-body model is detected to measure the overall inertial characteristics in multiple postures. an inertial property measurement unit;
a relative motion measurement unit for mounting the multi-body model under a pseudo free boundary condition, causing the multi-body model to move relative to the internal force between links, and measuring vibration or motion;
a control unit connected to the global inertial property measurement unit and the relative motion measurement unit, for coupling the global inertial properties and the vibrations or motions in the multiple postures to identify a minimum mechanical parameter of each of the objects; Equipped with a multi-body minimum dynamics parameter identification device. The control unit generates a first minimum dynamics parameter related to the mass of each object, a second minimum dynamics parameter related to the center of gravity of each object, and an inertia power of the inertia tensor of each object, based on the overall inertia characteristics at the multiple poses. 2. The multi-body minimum dynamics parameter identification device according to claim 1, wherein a third minimum dynamics parameter relating to the integral component is identified. wherein said control unit is further configured to identify a fourth minimum dynamics parameter for inertia diagonal term components of said inertia tensor of each said body based on said global inertia characteristics at said multiple poses and said relative motion of each said body. Item 3. The multi-body minimum dynamics parameter identification device according to item 2. Overall inertia characteristic measurement for measuring the overall inertia characteristics in multiple postures by detecting the natural angular frequency of free vibration of a multibody model in which multiple objects are connected under a pseudo free boundary condition process and
a relative motion measurement step of mounting the multi-body model under a pseudo free boundary condition, causing the multi-body model to move relative to the internal force between links, and measuring vibration or motion;
and a minimum dynamics parameter identification step for identifying minimum dynamics parameters of each of the objects by combining the overall inertial characteristics and the vibrations or motions in the multiple postures. The minimum mechanical parameter identification step includes:
a first minimum mechanical parameter identification step for identifying a first minimum mechanical parameter related to the mass of each of the objects based on the global inertial properties at the multiple poses;
a second minimum mechanical parameter identification step for identifying a second minimum mechanical parameter related to the center of gravity of each object based on the global inertial properties at the multiple poses;
and a third minimum dynamics parameter identification step for identifying a third minimum dynamics parameter related to the product of inertia components of the inertia tensor of each of the objects based on the global inertia characteristics at the multiple poses. A method for identifying the minimum mechanical parameters of the described multi-body system. The minimum dynamics parameter identifying step further identifies a fourth minimum dynamics parameter for the inertia diagonal term component of the inertia tensor of each of the bodies based on the overall inertia characteristics at the multiple poses and the relative motion of each of the bodies. 6. The multi-body minimum dynamics parameter identification method according to claim 5, comprising a fourth minimum dynamics parameter identification step for . Overall inertia characteristic measurement for measuring the overall inertia characteristics in multiple postures by detecting the natural angular frequency of free vibration of a multibody model in which multiple objects are connected under a pseudo free boundary condition a procedure;
A relative motion measurement procedure for mounting the multi-body model under a pseudo free peripheral boundary condition, causing the multi-body model to move relative to the internal force between links, and measuring vibration or motion;
A multi-body minimum mechanical parameter identification program comprising: a minimum mechanical parameter identification procedure for identifying a minimum mechanical parameter of each of the objects by coupling the overall inertial characteristics and the vibrations or motions in the multiple postures. The minimum mechanical parameter identification procedure includes:
a first minimum mechanical parameter identification procedure for identifying a first minimum mechanical parameter related to the mass of each of the objects based on the global inertial properties at the multiple poses;
a second minimum mechanical parameter identification procedure for identifying a second minimum mechanical parameter with respect to the center of gravity of each object based on the global inertial properties at the multiple poses;
and a third minimum dynamics parameter identification procedure for identifying a third minimum dynamics parameter related to the product of inertia components of the inertia tensor of each of the bodies based on the global inertia characteristics at the multiple poses. A minimal mechanical parameter identification program for the described many-body system. The minimum dynamics parameter identification procedure further identifies a fourth minimum dynamics parameter for the inertia diagonal term component of the inertia tensor of each of the bodies based on the global inertial properties at the multiple poses and the relative motion of each of the bodies. 9. The multi-body minimum dynamics parameter identification program according to claim 8, comprising a fourth minimum dynamics parameter identification procedure for .

Images