Body analysis system

Abstract

To enable a physical amount on a body movement that is difficult to measure to be analyzed from a physical amount on a body movement that is easy to measure.SOLUTION: A body analysis system (S) includes: measuring means (1) for measuring a physical amount on a body motion of a subject; simulation means (52) for simulating a motion of the subject, which simulates a state corresponding to a state of the body measured by the measuring means (1); and physical amount calculation means (54) for simulating the state of the body of the subject by the simulation means (52) using a physical amount on the body motion in which an error between a simulation result on the state of the body simulated by the simulation means (52) and a measurement result of the state of the body when measured by the measuring means (1), and calculating a physical amount on the body motion different from the measurement object in the measuring means (1).SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] This invention relates to a physical analysis system for analyzing the physical condition of a subject. [Background technology]

[0002] The technologies described in the following Patent Documents 1-3 are conventionally known as techniques for analyzing the physical condition and movements of a person (subject). Patent Document 1 (Reprint No. 2020 / 240749) describes a technique in which physical quantities related to body movement are measured multiple times in advance for each movement pattern (e.g., walking on level ground, walking uphill) to obtain sample data for learning (so-called training data), and walking patterns are analyzed using machine learning with a large amount of training data from information on body movement measured using an inertial sensor.

[0003] Patent Document 2 (Japanese Patent Publication No. 2021-083562) describes a technique for estimating external forces generated during body movement by measuring body movement data using a motion capture system and performing inverse kinematic analysis using a body model. In Patent Document 2, after defining a body model, the in-vivo load is calculated, the physical quantities that minimize the in-vivo load are calculated, and the optimal values ​​of the mechanical quantities are calculated.

[0004] Patent Document 3 (Japanese Patent Publication No. 2017-000546) describes a technique for reproducing walking motion measured using an inertial sensor with motion generation simulation using a neural oscillator, calculating a walking motion pattern from the simulation results, and estimating walking speed, stride length, etc., from the walking motion pattern. [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Re-tabled publication No. 2020 / 240749 ("0057" - "0071") [Patent Document 2] Japanese Patent Publication No. 2021-083562 ("0019"-"0026")

Patent Document 3

Summary of the Invention

Problems to be Solved by the Invention

[0006] (Problems of the Prior Art) In the technology described in Patent Document [1], a large amount of teacher data is required in terms of configuration. Therefore, there are problems that it takes time and computational load for the calculation of physical quantities, and the accuracy cannot be ensured when the data is scarce. In addition, there is also a problem that it cannot be applied to motion analysis where it is difficult to obtain teacher data, for example, motion analysis on ice or snow, or motion analysis targeting orthopedic disease patients or prosthetic users. In Patent Document [2], physical quantities are estimated by inverse kinematics, and there is a problem that the accuracy of the estimation result tends to decrease without accurate measurement data. For example, in order to estimate the force applied to the foot (ground reaction force) by the technology of Patent Document [2], if a large number of markers are not placed on the foot, a calculation result that protrudes from the foot may be calculated during the calculation by inverse kinematics, and there is also a risk that an unrealistic estimation result may be calculated.

[0007] In the motion generation simulation using neural oscillators described in Patent Document [3], since the neural oscillator is a mathematical model that generates periodic motions, there is a problem that Patent Document [3] can only be applied to periodic motions such as walking.

[0008] The technical problem of the present invention is to enable the analysis of physical quantities related to body movements that are difficult to measure from physical quantities related to body movements that are easy to measure.

Means for Solving the Problems

[0009] In order to solve the above technical problem, the body analysis system of the invention described in claim 1 a subject It is attached to a part of the body, and it is attached body part in motion It fluctuates accordingly. measurement means for measuring physical quantities, A simulation means for simulating the movements of the subject, wherein the measurement means measures When measuring the physical quantity based on the physical quantity A state corresponding to the physical condition By using a body model in which each segment of the body is composed of rigid links and rotational degrees of freedom are predefined at the joints of each rigid link, the joints are represented. The simulation means for simulating, The simulation results of the physical state simulated by the simulation means, and the measurement results of the physical state measured by the measurement means, Based on this, the value of the evaluation function for evaluating body motion generated by forward dynamics simulation was calculated to minimize the value. Using physical quantities, the simulation to By using this method, the physical condition of the subject is simulated, ground reaction force of From the simulation results and the body model Means for calculating the physical quantity to be calculated and , The physical quantity calculation means simulates the movement of the subject using a contact model of the subject's foot and the contact surface, It is characterized by having the following features.

[0010] The invention described in claim 2 is a body analysis system described in claim 1, Based on the information of the angle and angular velocity of the subject's joints, the subject's joint torque is derived, and based on the derived joint torque, the measurement target of the measurement means is determined. Equivalent to A simulation means for simulating physical quantities related to body movement, namely the angle and angular velocity of a subject's joints, by deriving them using forward dynamics, wherein the simulation means feeds back the derived joint angle and angular velocity into the derivation of the joint torque for further simulation. It is characterized by having the following features. [Effects of the Invention]

[0012] According to the invention described in claim 1, physical quantities relating to physical motion that are difficult to measure can be analyzed from physical quantities relating to physical motion that are easy to measure. Furthermore, according to the invention described in claim 1, the movements of the subject can be simulated in a situation where the subject's feet are in contact with the floor surface. According to the invention described in claim 2, the accuracy of the calculation results can be improved compared to the case where forward dynamics is not used. Furthermore, the accuracy of the calculation results can be improved compared to the case where no feedback is provided. 。 [Brief explanation of the drawing]

[0013] [Figure 1] Figure 1 is an explanatory diagram of the body analysis system according to Embodiment 1 of the present invention. [Figure 2] Figure 2 is a functional block diagram of the main unit of the information processing device in the body analysis system of Example 1. [Figure 3] Figure 3 is an explanatory diagram of an example of a body model used in the body analysis system of Example 1. [Figure 4] Figure 4 is a schematic diagram illustrating the simulation of Example 1. [Figure 5] Figure 5 is an explanatory diagram of the sensor model. Figure 5A shows the subject standing, Figure 5B shows the subject sitting, Figure 5C shows the subject lying face down, and Figure 5D shows the subject standing with their upper limb (right hand) raised horizontally. [Figure 6] Figure 6 is an explanatory diagram of the main parts of the contact point with the floor in the body model of Example 1. [Figure 7] Figure 7 is an explanatory flowchart of the body analysis system in Example 1. [Modes for carrying out the invention]

[0014] Next, specific examples of embodiments of the present invention (hereinafter referred to as "examples") will be described with reference to the drawings, but the present invention is not limited to the following examples. In the following explanation using diagrams, diagrams of components other than those necessary for the explanation have been omitted as appropriate for ease of understanding. [Examples]

[0015] Figure 1 is an explanatory diagram of the body analysis system according to Embodiment 1 of the present invention. In Figure 1, the body analysis system S of Example 1 has a sensor unit 1 as an example of a measurement means to be attached to the subject. The sensor unit 1 of Example 1 has an inertial sensor built in. The inertial sensor of Example 1 is a sensor that can detect acceleration and angular velocity, and as an example, a configuration combining a 3-axis angular velocity meter and a 3-axis accelerometer can be adopted. In Example 1, the sensor unit 1 is attached to the subject's toes (1a), foot (1b), lower leg (1c) or thigh (1d), forearm (1e), pelvis (1f), and trunk (1g). The attachment position of the sensor unit 1 is not limited to the exemplified positions, and it can be attached to any position, and the number can also be increased or decreased. Each sensor unit 1 (1a to 1g) in Example 1 incorporates a wireless communication module as an example of a communication component. In Example 1, a wireless LAN module is used as an example of a wireless communication module, but the system is not limited to this, and any communication module, such as a mobile phone network or Bluetooth®, can be used. Furthermore, the body analysis system S of Embodiment 1 includes a personal computer 11 as an example of an information processing device. The personal computer 11 includes a computer body 12, a display 13 as an example of a display unit, and a keyboard 14 and a mouse 15 as an example of an input unit. The body 12 has a wireless communication module (not shown) built in, and is configured to enable wireless communication with the sensor unit 1.

[0016] (Description of the control unit of the computer main unit 12 in Example 1) Figure 2 is a functional block diagram of the main unit of the information processing device in the body analysis system of Example 1. In Figure 2, the control unit 41 of the computer main unit 12 in Embodiment 1 is composed of a computer device having an I / O (input / output interface) that performs input / output of signals to and from the outside and adjustment of input / output signal levels, a ROM (read-only memory) that stores programs and data for performing necessary startup processing, a RAM (random access memory) for temporarily storing necessary data and programs, a CPU (central processing unit) that performs processing according to the startup program stored in the ROM, etc., and a clock oscillator, etc. Various functions can be realized by executing the programs stored in the ROM and RAM, etc. The control unit 41 stores basic software that controls basic operations, so-called operating system OS, a body analysis program AP1 as an example of an application program, and other software not shown.

[0017] (Element connected to the control unit 41 of Example 1) The control unit 41 receives output signals from signal output elements such as the keyboard 14, mouse 15, and sensor unit 1 (1a to 1g). Furthermore, the control unit 41 of Embodiment 1 outputs control signals to controlled elements such as the display 13.

[0018] (Functions of the control unit 41) The body analysis program AP1 of the control unit 41 in Example 1 has the following functional means (program modules) 51 to 54.

[0019] The measurement data acquisition means 51 acquires physical quantities related to the subject's physical movements measured by each sensor unit 1a to 1g. In Example 1, measurement data of angular velocity and acceleration measured by each sensor unit 1a to 1g are acquired as physical quantities related to the subject's physical movements. The simulation means 52 simulates (mimics) the actions of the subject and simulates a state corresponding to the physical state measured by the sensor units 1a to 1g. The simulation means 52 of Embodiment 1 has the following means 52a to 52g.

[0020] Figure 3 is an explanatory diagram of an example of a body model used in the body analysis system of Example 1. The body model memory means 52a of the simulation means 52 stores the body model to be simulated. In Figure 3, the body model 101 of Embodiment 1 is composed of rigid links for each segment of the body, and rotational degrees of freedom are defined at the joints of each ring to represent joints. In Embodiment 1, as an example, the total number of links for the whole body is 8 links for both legs (4 links for each leg: thigh, lower leg, foot, and toes), 2 links for the trunk and pelvis, and 4 links for both arms (2 links for each arm: upper arm and forearm), for a total of 14 links. The rotational degrees of freedom for each joint are as follows: the hip joint has 3 degrees of freedom, the knee joint has 1 degree of freedom (flexion, extension, and straightening), the ankle joint has 1 degree of freedom (plantarflexion and dorsiflexion), the toe joint has 1 degree of freedom (plantarflexion and dorsiflexion), the lumbar joint has 3 degrees of freedom, the shoulder joint has 3 degrees of freedom, and the elbow joint has 1 degree of freedom (flexion, extension, and straightening), for a total of 23 degrees of freedom. Each joint angle is defined as 0° when the body is in a static standing position as shown in Figure 3. The link length, moment of inertia, and center of mass position of each link are determined using estimation formulas based on the subject's height and weight. Since various previously known estimation formulas, such as those described in "Ae, M. et al, Estimation of inertia properties of the body segments in Japanese athletes, Society of Biomechanism Japan, Vol.11 (1992), pp.23-33" and "Drillis, R. et al, Body segment parameter: A survey of measurement techniques, Artificial Limbs, Vol.8, No.1 (1964), pp.329-351," can be used, a detailed explanation will be omitted.

[0021] The spline interpolation processing means 52b performs the simulation on N node points at time t, i.e., at joint j, from time t1 to t N Coordinate data for N time history up to (q j) each (q1 j ,t1),(q2 j ,t2),…,(q N j ,t N Based on this, the target joint angle is interpolated using spline interpolation. TIFF0007911388000001.tif1328 and target joint angular velocity Calculate TIFF0007911388000002.tif1329. In Example 1, the target joint angle TIFF0007911388000003.tif1430 is a cubic spline interpolation method defined by the following equation 1.

number

number

[0022] The PD control means 52c controls the target joint angle TIFF0007911388000008.tif1329 and target joint angular velocity Based on TIFF0007911388000009.tif1430, the operating torque for tracking the target attitude. TIFF0007911388000010.tif1321 is calculated by PD (Proportional-Differential) control. The PD control means 52c of Example 1 is the operating torque TIFF0007911388000011.tif1321 was defined by Equation 3 below.

Equation

[0023] The muscle driving torque calculation means 52d calculates the muscle driving torque based on TIFF0007911388000013.tif1321. The muscle driving torque calculation means 52d in Example 1 defined the muscle driving torque TIFF is 0007911388000014.tif1530. TIFF0007911388000015.tif1431 was defined by Equation 4 below.

Equation

[0024] TIFF0007911388000018.tif1419 represents the normalized muscle strength of the muscles that contribute to torque generation at joint j. Normalized muscle strength TIFF0007911388000019.tif1418 was calculated based on the relationship between muscle strength, muscle activity, muscle contraction velocity, and muscle length, according to the Hill-type model for contraction dynamics of muscle tissue. The relationship is described in "Thelen, DG, Adjustment of muscle mechanics model parameters to simulate dynamic contractions in older adults, Journal of Biomechanical Engineering, Vol.125 (2003), pp.70-77," so a detailed explanation is omitted. In Example 1, since muscle fibers were not modeled, muscle activity m j (t), normalized muscle contraction velocity TIFF0007911388000020.tif1319, normalized muscle length TIFF0007911388000021.tif1421 was defined using equations 5, 6, and 7 based on joint torque, joint angle, and joint angular velocity.

number

number

number

[0025] The resistance torque calculation means 52e calculates the resistance torque generated by the passive resistance of each joint. Calculate TIFF0007911388000029.tif1434. That is, the resistive torque. TIFF0007911388000030.tif1432 corresponds to reproducing the range of motion of a joint. The resistance torque calculation means 52e of Example 1 calculates the resistance torque in the following number 8. Define TIFF0007911388000031.tif1432.

number

[0026] The joint torque calculation means (feedback means) 52f calculates the muscle-driven torque generated by muscle force. TIFF0007911388000038.tif1330 and the resistive torque generated by the passive resistance of the joint Based on TIFF0007911388000039.tif1433, joint torque Calculate TIFF0007911388000040.tif1626. Joint torque TIFF0007911388000041.tif1526 is calculated from the following number 9.

number

[0027] The forward dynamics calculation means 52g calculates the joint torque at joint j during the simulation time t. Joint movement when TIFF0007911388000043.tif1427 is applied The TIFF0007911388000044.tif1254 is calculated using forward dynamics calculations. In Example 1, the joint angles derived from the forward dynamics calculations are used. TIFF0007911388000045.tif1217, angular velocity TIFF0007911388000046.tif1217, angular acceleration TIFF0007911388000047.tif1218 is used in calculations by the PD control means 52c, muscle drive torque calculation means 52d, and resistance torque calculation means 52e when iterative calculations are performed, thereby generating joint torque. This is fed back into the calculation of TIFF0007911388000048.tif1526.

[0028] Figure 4 is a schematic diagram illustrating the simulation of Example 1. Therefore, in the simulation means 52 of Example 1, as shown in Figure 4, N node points (q1 j ,t1),(q2 j ,t2),…,(q N j ,t N From the input of ), the spline interpolation means 52b processes the target joint angle TIFF0007911388000049.tif1430 and target joint angular velocity TIFF0007911388000050.tif1230 is calculated. Then, the target joint angle is controlled by the PD control means 52c. TIFF0007911388000051.tif1331 and target joint angular velocity Operating torque from TIFF0007911388000052.tif1329 TIFF0007911388000053.tif1221 is calculated, and in the muscle drive torque calculation means 52d, the operating torque Muscle-driven torque from TIFF0007911388000054.tif1320 TIFF0007911388000055.tif1431 is calculated. Then, the joint torque calculation means 52f calculates the muscle drive torque. TIFF0007911388000056.tif1330 and resistance torque Joint torque from TIFF0007911388000057.tif1432 TIFF0007911388000058.tif1327 is calculated, and the joint torque is calculated by the forward dynamics calculation means 52g. Physical quantities of joint movement from TIFF0007911388000059.tif1526 The file TIFF0007911388000060.tif1153 is calculated. In Example 1, the simulation means 52 constructs a digital body model 101 using MATLAB® / Simulink® based on the definitions of each body segment and joint of the body model 101, and applies joint torque to each joint (j) of the body model 101. Entering TIFF0007911388000061.tif1230 will enable joint motion (angular acceleration). TIFF0007911388000062.tif1218, angular velocity TIFF0007911388000063.tif1118, angle A forward dynamics simulation was constructed to calculate TIFF0007911388000064.tif1218). The variable-step solver ode23 was used for the simulation.

[0029] The minimum error physical quantity calculation means 53 calculates the physical quantity related to the body movement that minimizes the error between the simulation result of the body state simulated by the forward dynamics simulation (simulation means 52) and the measurement result of the body state measured by the sensor units 1a to 1g. The minimum error physical quantity calculation means 53 of Example 1 sets an evaluation function for evaluating the body movement generated by the forward dynamics simulation and calculates the physical quantity of joint movement that minimizes the value of the evaluation function. The value TIFF0007911388000065.tif1263 is derived through optimization calculation. The optimization calculation uses a genetic algorithm (GA). Note that the optimization calculation using a genetic algorithm is publicly known, as described in works such as "Holland, JH, Adaptation in natural and artificial systems: An introductory analysis with applications to biology, control, and artificial intelligence, The MIT press (1992)," so a detailed explanation is omitted.

[0030] Joint movement The parameters necessary to generate TIFF0007911388000066.tif1264 are set as variables, and the variable values ​​that minimize the value of the evaluation function defined in the following equation (number 10) are derived.

number

number

number

number

[0031] (Description of the sensor model) Figure 5 is an explanatory diagram of the sensor model. Figure 5A shows the subject standing, Figure 5B shows the subject sitting, Figure 5C shows the subject lying face down, and Figure 5D shows the subject standing with their upper limb (right hand) raised horizontally. To obtain values ​​equivalent to the measured values ​​of the inertial sensor (sensor unit) 1 from the results of forward dynamics simulation, a sensor coordinate system was defined on each body segment of the digital human model. This allows the system to calculate values ​​equivalent to the measured values ​​of the inertial sensor from the simulation results by outputting values ​​representing the relative acceleration and angular velocity between the spatial coordinate system and the sensor coordinate system in the sensor coordinate system, based on the joint movements calculated by forward dynamics calculations. In Example 1, sensor coordinate systems were defined at a total of seven locations on the trunk, pelvis, forearm, thigh, lower leg, foot, and toe segments. The orientation of each sensor coordinate system was calibrated using the method described below. The method in Example 1 performs calibration by calculating the orientation of the sensor coordinate system relative to the spatial coordinate system using the stationary measured values ​​from the inertial sensor 1.

[0032] First, as shown in Figure 5A, when standing with the body facing forward towards the Y-axis of the spatial coordinate system, the rotation matrix representing the orientation of the sensor coordinate system relative to the spatial coordinate system is used. Define TIFF0007911388000075.tif1226 using the ZYX Euler angles in equation 14.

number

number

[0033] From number 15, ψ s and θ s This is calculated using numbers 16 and 17.

number

number

number

[0034] Furthermore, consider the posture shown in Figure 5C, where the upper body is tilted 90° forward from the posture in Figure 5A. In this case, the posture of the inertial sensor 1 (1e~1g) attached to the upper body is rotated -90° around the X-axis relative to the posture in Figure 5A. Therefore, the rotation matrix representing the posture of the upper body's sensor coordinate system relative to the spatial coordinate system in the posture of Figure 5C. The number TIFF0007911388000083.tif1234 is 19.

number

[0035] Furthermore, consider the posture shown in Figure 5D, where the shoulder joint is abducted by 90° from the posture in Figure 5A. In this case, the posture of the inertial sensor 1e attached to the upper limb is rotated -90° around the Y-axis relative to the posture in Figure 5A. Therefore, the rotation matrix representing the posture of the upper limb's sensor coordinate system relative to the spatial coordinate system in the posture of Figure 5D. TIFF0007911388000085.tif1232 is the number 20.

number

[0036] From equations 18 to 20, the relationship between the measurement value from inertial sensor 1 and gravitational acceleration is given by equations 21, 22, and 23.

number

number

number

[0037] Here, TIFF0007911388000090.tif1243 is the measured acceleration value of inertial sensor 1 in the posture shown in Figure 5B. TIFF0007911388000091.tif1141 is the measured acceleration of the inertial sensor in the posture shown in Figure 5C. TIFF0007911388000092.tif1244 is the measured acceleration of the inertial sensor in the attitude shown in Figure 5D. From equations 21 to 23, ψ s These values ​​are calculated using equations 24, 25, and 26 for the inertial sensors of the lower limbs, upper body, and upper limbs, respectively.

number

number

number

[0038] In Example 1, the minimum error physical quantity calculation means 53 uses the calculated variables to perform a resimulation with the simulation means 52, calculates the evaluation function from the results of the resimulation, and recalculates the variables. This process is repeated until the evaluation function is minimized or the calculation of the variables is repeated a predetermined number of times, at which point the calculation of the variables is terminated.

[0039] The physical quantity calculation means 54 uses the physical quantities related to body movements that minimize errors (optimized) to simulate the state of the subject's body in the simulation means 52 and calculates physical quantities related to body movements different from those measured by the sensor unit 1. In Example 1, the physical quantity calculation means 54 uses the variables derived by the minimum error physical quantity calculation means 53 to perform a simulation in the simulation means 52. From the simulation results, it calculates the ground reaction force, which is an external force acting on the body from the ground, as an example of a physical quantity related to body movements different from those measured by the sensor unit 1. The physical quantities related to body movements different from those measured by the sensor unit 1 can also be applied to, for example, joint angles and joint moments, and are particularly suitable for physical quantities that are difficult to measure directly or require special equipment.

[0040] In Example 1, the physical quantity calculation means 54 defined the ground reaction force acting on the body model 101 from the floor surface as an external force acting on the feet of the body model 101 during the simulation. Vertical ground reaction force TIFF0007911388000096.tif1318 was defined by equation 27 based on the amount and velocity of penetration of the contact point into the floor surface (Hwang et al., 2003).

number

[0041] Next, the horizontal reaction force TIFF0007911388000101.tif1641 is defined using equations 28 and 29, based on the same method as for vertical ground reaction forces.

number

number

[0042] Let μ be the coefficient of friction between the contact point and the floor surface. TIFF0007911388000107.tif1524 or TIFF0007911388000108.tif1422 is When the value exceeds TIFF0007911388000109.tif1525, slippage occurs between the contact point and the floor surface. Therefore, horizontal floor reaction force in a stationary state or when slippage is occurring. TIFF0007911388000110.tif1535 was defined using numbers 30 and 31.

number

number

[0043] Figure 6 is an explanatory diagram of the main parts of the contact point with the floor in the body model of Example 1. Figure 6 shows the positions of the contact points of the body model in Example 1. In Figure 6, the contact points were set at four points on the foot segment (medial and lateral heel, medial and lateral midfoot) and two points on the toe segment (medial and lateral toe). In addition, a virtual torque τ was set to constrain the rotation of the body model in order to prevent the body model from falling over during the simulation. virtual (t) was defined. A virtual joint was set between the center of mass position of the pelvic segment of the body model and the spatial coordinate system, and a virtual torque τ acting on the virtual joint was defined. virtual (t) was defined in equation 32.

number

[0044] (Explanation of the flowchart for Example 1) Next, the control flow in the body analysis system of Example 1 will be explained using a flowchart.

[0045] (Explanation of the flowchart) Figure 7 is an explanatory flowchart of the body analysis system in Example 1. Each step ST in the flowchart of Figure 7 is performed according to a program stored in the computer unit 12. This process is executed in parallel with various other processes in the computer unit 12. The flowchart shown in Figure 7 is initiated when the body analysis program AP1 is started on the computer unit 12.

[0046] In ST1 of Figure 7, the initial variables (assumed values) of the simulation are input to the simulation means 52. Then, the process proceeds to ST2. In ST2, the simulation means 52 generates body movements based on the input variables. Then, the process proceeds to ST3. In ST3, measurement data is acquired from sensor unit 1. Then, the process proceeds to ST4. In ST4, calculate the evaluation function from the simulation results and measurement results (measurement data), and then proceed to ST5. In ST5, determine whether the evaluation function has been minimized. If yes (Y), proceed to ST6; if no (N), proceed to ST7. In ST6, the optimal variables for the simulation are updated, and the process proceeds to ST7. In ST7, it is determined whether the evaluation function calculation has been performed the specified number of times. If yes (Y), proceed to ST9; if no (N), proceed to ST8. In ST8, the variables input to the simulation means 52 are updated, and the program returns to ST2 to perform the simulation again. In ST9, the simulation means 52 uses the optimal variables to perform a simulation and generate body motion. Then, the process proceeds to ST10. In ST10, the simulation results are output. Then, the process proceeds to ST11. In ST11, physical quantities (such as ground reaction forces) are calculated (estimated) based on the simulation results output in ST10. The simulation results and the calculated ground reaction forces are output to display 13. Then, the body analysis program AP1 is terminated.

[0047] (Effect of Example 1) In the body analysis system S of Embodiment 1, which has the above configuration, physical quantities related to body movement are calculated to minimize the error between the actual measurement results and the simulation results (so-called optimized physical quantities), and unmeasured physical quantities are calculated using the simulation results that utilize these optimized physical quantities. Therefore, even physical quantities that are difficult to measure in reality, or that require expensive or special equipment or facilities for measurement, can be estimated from the simulation results. For example, measuring ground reaction force requires a force plate and is also restricted in the way the subject walks, making measurement cumbersome. However, in Embodiment 1, ground reaction force can be estimated from measurement data when the subject walks, without using a special device such as a force plate. Therefore, in Example 1, a large amount of training data like that in Patent Document 1 is not required, and the degradation seen in Patent Document 2, which utilizes inverse kinematics, is suppressed. Furthermore, unlike Patent Document 3, which is limited to walking, Example 1 can be used as an estimation method for activities other than walking.

[0048] (Application of Example 1) By modifying the external force model in Example 1, it is possible to apply it to movements other than walking on level ground. Here, we will explain using skating as an example. Vertical ground reaction force acting on the foot (skate) of body model 101. TIFF0007911388000115.tif1217 is defined by equation 33, based on the amount and velocity of the contact point's penetration into the ice rink, similar to when walking on flat ground (see equation 27).

number

number

number

[0049] The coefficient of friction between the contact point of the blade in the horizontal direction and the ice rink is μ h The coefficient of friction between the contact point perpendicular to the blade and the ice rink is μ. v The horizontal ground reaction force of the blade in a stationary or sliding state. TIFF0007911388000130.tif1217 and blade vertical ground reaction force Define TIFF0007911388000131.tif1218 using numbers 36 and 37, similar to numbers 30 and 31.

number

number

[0050] (Example of change) Although embodiments of the present invention have been described in detail above, the present invention is not limited to the embodiments described above, and various modifications can be made within the scope of the gist of the present invention as described in the claims. Examples of modifications to the present invention (H01) to (H04) are shown below. (H01) In the above embodiment, a configuration in which processing is performed on a single personal computer 11 was illustrated, but the invention is not limited thereto. For example, it is also possible to perform distributed processing on multiple personal computers 11 or servers connected to a network.

[0051] (H02) In the above embodiment, the specific numerical values ​​exemplified can be changed as appropriate depending on the design and specifications. For example, the number of inertial sensors 1 can be set arbitrarily. That is, in the method of the embodiment, it is possible to estimate the ground reaction force with as few as one sensor, but by using more sensors than the number shown in the embodiment, it is possible to improve the accuracy of the estimation results. Conversely, by using fewer sensors than the number shown in the embodiment, it is possible to perform measurement and processing in a simpler and faster manner.

[0052] (H03) In the above embodiment, an example was given of a configuration in which physical quantities related to body movement are measured using an inertial sensor, but the embodiment is not limited thereto. Any system capable of measuring physical quantities related to body movement, such as a motion capture system, can be used. When using a motion capture system, the number of markers can also be set arbitrarily, as with the inertial sensor 1, as explained in (H02). In other words, the embodiment is not limited to a configuration in which acceleration and angular velocity are measured by the inertial sensor 1 as physical quantities related to body movement, but can also be used with a motion capture system that measures the position of markers as a physical quantity related to body movement. Furthermore, even when using a motion capture system, it is also possible to use a method that measures body position from image information without using markers (markerless motion capture system).

[0053] (H04) In the above embodiment, the external force models were exemplified as the case of floor and foot, and ice and ice skates, but are not limited to these. For example, by using an external force model of snow and foot, it is possible to estimate the ground reaction force on snow, or by using an external force model of floor-prosthesis, it is possible to estimate the ground reaction force for a prosthesis user. [Explanation of Symbols]

[0054] 1... Measuring means, inertial sensor, 52…Means of simulation, 54...Physical quantity calculation means, S...Body analysis system.

Claims

1. A measuring means that is attached to a part of the subject's body and measures a physical quantity that changes in accordance with the movement of the body part to which it is attached, A simulation means for simulating the movements of the subject, wherein the simulation means simulates a state corresponding to the state of the body at the time of measurement of a physical quantity measured by the measurement means, using a body model in which each segment of the body is composed of rigid links and rotational degrees of freedom are predetermined at the joints of each rigid link to represent joints, A physical quantity calculation means that simulates the state of a subject's body using the simulation means, and calculates the ground reaction force from the simulation results and the body model, using a physical quantity calculated to minimize the value of an evaluation function for evaluating body motion generated by forward dynamics simulation, based on the simulation results of the state of the body simulated by the simulation means and the measurement results of the state of the body measured by the measurement means, the physical quantity calculation means that simulates the movement of the subject using a contact model of the subject's foot and the contact surface, A body analysis system characterized by having the following features.

2. A simulation means that derives the joint torque of the subject based on information of the angle and angular velocity of the joints of the subject's body, and simulates by deriving the angle and angular velocity of the subject's joints, which are physical quantities relating to the body movement corresponding to the object measured by the measurement means, using forward dynamics based on the derived joint torque, wherein the simulation means feeds back the derived joint angle and angular velocity into the derivation of the joint torque during the simulation. The body analysis system according to claim 1, characterized by comprising the following features.

Images