JP5991532B2 - Human motion evaluation apparatus, method, and program - Google Patents

Description

  The present invention relates to a technique for evaluating the movement of a human body, and more particularly to a technique for determining a reaching movement that increases positioning accuracy for a human using a musculoskeletal model.

  Ergonomics, which attempts to use human physical characteristics and physiological characteristics in engineering, is actively applied to the development of human interfaces for various industrial products. Ergonomic design is not only easy for humans to use, but also helps to prevent mistakes that humans are likely to make.

  In ergonomics, the human body is represented by various models, and various movements of the human body are simulated on the computer. As an example of the model, there is a finite element model as the most complicated one, and a musculoskeletal model obtained by modeling a human skeleton, joint, and skeletal muscle as a simple one. For example, as an evaluation of human movement based on a musculoskeletal model, an evaluation system has been proposed by the inventors of the present application that includes an evaluation index that quantitatively evaluates skill and sensibility and can automatically calculate a posture close to a human posture. (For example, refer to Patent Document 1).

JP 2011-141706 A

  In the conventional human body movement evaluation based on the musculoskeletal model, the time integral value of the muscle activity of each muscle when the musculoskeletal model is subjected to various reaching movements and its sum are calculated, and the sum of the time integral values is small. It is evaluated that the movement efficiency of the reaching movement is high. Based on such a minimum energy consumption criterion, there is an advantage that it is possible to determine a movement that is less likely to cause a person to get tired, that is, a movement with excellent energy efficiency.

  On the other hand, when the brain commands exercise to skeletal muscles, it is known that noise corresponding to exercise intensity (motion command dependent noise) is superimposed on the command. Based on this knowledge, a minimum end point variance criterion has been proposed in which human beings select a motion trajectory that minimizes motion command-dependent noise. Based on the minimum endpoint variance norm, it is expected that humans can evaluate natural and easy-to-move movements that could not be evaluated with the minimum energy consumption norm.

  In view of the above problems, an object of the present invention is to make it possible to evaluate the movement of a human body from the viewpoint of positioning accuracy.

  A human body motion evaluation device according to one aspect of the present invention is a device that evaluates human body motion using a musculoskeletal model obtained by modeling the human skeleton, joints, and skeletal muscles, and the limb of the musculoskeletal model is A joint angle calculation unit that receives a motion trajectory of the limb when reaching a reaching motion and calculates a temporal angle change of each joint in the musculoskeletal model; A joint torque calculation unit for calculating a typical torque change, a muscle force calculation unit for calculating a change in strength of each skeletal muscle over time in the musculoskeletal model from a change in torque over time of each joint, A muscle activity calculation unit that calculates a change in muscle activity over time of each skeletal muscle from a change in muscle strength over time, and a motion command-dependent noise superimposed on a change in muscle activity over time of each skeletal muscle The limb performs the reaching movement An end point variation evaluating unit that evaluates the variation in the reaching end point of the mushroom, and various variations in the reaching motion of the limb, and based on the variation in the reaching end point of the reaching motion evaluated by the end point variation evaluating unit for each reaching motion For this, a reaching motion determination unit for determining a reaching motion that increases positioning accuracy is provided.

  A human body motion evaluation method according to one aspect of the present invention is a method for evaluating human body motion using a musculoskeletal model in which the skeleton, joints, and skeletal muscles of a human body are modeled. Receiving the movement trajectory of the limb when the limb makes a reaching movement, calculate the temporal change in angle of each joint in the musculoskeletal model, and calculate the change in torque of each joint over time from the change in angle of each joint Then, a change in muscle strength of each skeletal muscle over time in the musculoskeletal model is calculated from a change in torque of each joint over time, and a time-dependent muscle activity of each skeletal muscle is calculated from a change in muscle strength of each skeletal muscle over time. A degree change is calculated, and a variation in reaching end point when the limb performs the reaching movement in a state where movement command-dependent noise is superimposed on a change in muscle activity over time of each skeletal muscle, Change the reaching movement of Based on the variation of the reaching movements endpoint was the evaluation for each reaching movement determines reaching movements to increase the positioning accuracy to humans.

  The human body motion evaluation program according to one aspect of the present invention is a program for causing a computer to evaluate human body motion using a musculoskeletal model obtained by modeling the human skeleton, joints, and skeletal muscles. A joint angle calculating means for calculating a temporal change in angle of each joint in the musculoskeletal model in response to a movement trajectory of the limb when the limb of the model performs a reaching movement; A joint torque calculating means for calculating a change in torque of a joint over time, a muscle force calculating means for calculating a change in strength of each skeletal muscle over time in the musculoskeletal model from a change in torque of each joint over time, and each skeletal muscle Muscle activity calculation means for calculating a change in muscle activity over time of each skeletal muscle from a change in muscle strength over time, and movement command dependent noise is superimposed on the change in muscle activity over time of each skeletal muscle The end point variation evaluating means for evaluating the end point variation when the limb performs the reaching movement in the state, and the reaching movement of the limb was changed variously, and each reaching movement was evaluated by the end point variation evaluating means. The computer is caused to function as a reaching motion determination means for determining a reaching motion that increases positioning accuracy for humans based on variations in the reaching motion end point.

  According to these human body motion evaluation apparatuses, methods, and programs, variations in reaching movement end points are evaluated for various reaching movements of the limb based on the musculoskeletal model, and positioning is performed for humans based on the evaluated variations in reaching movement end points. A reaching movement that increases accuracy is determined. The limb refers to the upper arm, the lower limb, the hand tip, the fingertip, the foot tip, and the like. Thereby, the motion of the human body can be evaluated from the viewpoint of positioning accuracy.

  In each of the human body motion evaluation apparatus, method, and program, the reaching motion is repeated on the limb a plurality of times by superimposing the motion command-dependent noise on the muscular activity change of each skeletal muscle over time, and each time. The reaching movement end point is calculated, the error between each reaching movement end point and the reference end point of the reaching movement is calculated, and the average value of the errors can be calculated to evaluate the variation of the reaching movement end point. .

  Alternatively, in each of the human body motion evaluation apparatus, method, and program, a time integral value of variance defined for the motion command dependent noise is calculated for each skeletal muscle in the musculoskeletal model, and the sum of these time integral values is calculated. By calculating, it is possible to evaluate the variation of the reaching movement end point.

  According to the present invention, the motion of the human body can be evaluated from the viewpoint of positioning accuracy.

Functional block diagram of the main part of the human body motion evaluation apparatus according to an embodiment of the present invention Schematic diagram of a musculoskeletal model (two-joint, six-muscle upper limb model) according to an example The figure which shows the exercise | movement trajectory of a hand and the muscular strength change of each skeletal muscle when making the reaching movement in the 2 joint 6-muscle upper limb model shown in FIG. The figure which shows the dispersion | variation in the end point of a certain reaching | attainment motion (motion trajectory natural for humans) when the motion command dependence noise is superimposed The figure which shows the dispersion | variation in the end point of another reaching motion (motion trajectory unnatural for humans) when motion command dependence noise is superimposed Schematic diagram showing the reaching movement to be judged by the reaching movement judging unit The figure which shows the dispersion | variation in the end point of each reaching movement shown in FIG. Box-and-whisker diagram showing the average error value of the end points of each reaching movement shown in FIG. Schematic diagram showing the outline of the experiment of human reaching movement The figure which shows the dispersion | variation in the end point of each reaching movement shown in FIG. Box-and-whisker diagram showing the average error value of the end points of each reaching movement shown in FIG.

  DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings. In addition, this invention is not limited to the following embodiment.

  FIG. 1 is a functional block diagram of the main part of the human body motion evaluation apparatus 10 according to the present embodiment. The human body motion evaluation apparatus 10 includes a joint angle calculation unit 11, a joint torque calculation unit 12, a muscle force calculation unit 13, a muscle activity level calculation unit 14, an end point variation evaluation unit 15, and a reaching motion determination unit 16. Note that the human body motion evaluation apparatus 10 can be implemented as dedicated hardware in which each of the above-described components is configured by a semiconductor integrated circuit or the like. Alternatively, the human body motion evaluation apparatus 10 can be implemented on a general-purpose computer by describing each of the above-described components by a computer program and causing a general-purpose computer such as a PC to execute the computer program. Furthermore, the human body motion evaluation apparatus 10 can be implemented by combining hardware and software.

  The human body motion evaluation apparatus 10 uses a musculoskeletal model to determine a reaching motion that increases positioning accuracy for a human. For convenience of explanation, the present embodiment uses a bi-joint 6-muscle upper limb model as shown in FIG. 2 as the musculoskeletal model, but the musculoskeletal model usable in the present invention is not limited to this.

The joint angle calculation unit 11 receives the motion trajectory of the hand when the hand (limb) of the two-joint six-muscle upper limb model (musculoskeletal model) makes a reaching movement, and the joint angle calculation unit 11 changes over time of each joint in the two-joint six-muscle upper limb model. To calculate the angle change. In the 2-joint 6-muscle upper limb model, the hand moves on a plane, and therefore, the coordinate position (X, Y) of the hand point 101 is input to the joint angle calculation unit 11 as the reaching movement of the hand. Then, the joint angle calculation unit 11 calculates the angles q ₁ and q ₂ of the joints J ₁ and J ₂ shown in FIG. 2 from the input coordinate position (X, Y). The joint angles q ₁ and q ₂ are parameters that change over time according to changes in the coordinate position (X, Y) of the hand point 101.

  The coordinate position (X, Y) of the hand point 101 may be numerically input to the human body motion evaluation apparatus 10, or a motion capture (not shown) attached to the human body is connected to the human body motion evaluation apparatus 10. It can also be obtained from the motion capture.

Joint torque calculation unit 12 calculates the temporal change in torque of the joint _J 1, _{J 2} from the temporal angular change of the joints _J 1, _{J 2} (joint angle _q 1, _{q 2).} The joint torque calculation unit 12 calculates the torques t ₁ and t ₂ of the joints J ₁ and J ₂ shown in FIG. 2 based on, for example, an equation of motion when the hand point 101 moves with the external force F. The joint torques t ₁ and t ₂ are parameters that change over time according to changes in the coordinate position (X, Y) of the hand point 101.

The muscular strength calculation unit 13 calculates the muscular strength change with time of each skeletal muscle in the musculoskeletal model from the temporal torque changes (joint torques t ₁ , t ₂ ) of the joints J ₁ and J ₂ . As shown in FIG. 2, the two-joint six-muscle upper limb model has six muscle strengths F _{1 to} F ₆ , and the muscle strength calculation unit 13 generates muscle strength as a motion of each skeletal muscle that generates joint torques t ₁ and t _2. F _{1 to} F ₆ are calculated. The muscular strengths F _{1 to} F ₆ are parameters that change over time according to changes in the coordinate position (X, Y) of the hand point 101.

FIG. 3 shows the movement trajectory of the hand and the muscular strength change of each skeletal muscle when the reaching movement in the two-joint six-muscle upper limb model shown in FIG. 2 is performed. FIG. 3A is a diagram in which a 2-joint 6-muscle upper limb model is projected on the xy plane. Joints _{J 1} is fixed at the origin of the xy plane, the hand points 101 to the reaching movement from the start position along a movement path 102 (0, 0.2) to the end position (0.3,0.2) . FIG. 3B is a graph obtained by plotting the muscle strengths F _{1 to} F ₆ calculated by the muscle strength calculation unit 13 when the reaching motion is performed on the bi-joint 6-muscle upper limb model.

Returning to FIG. 1, the muscle activity calculation unit 14 calculates the change in muscle activity with time of each skeletal muscle from the change in muscle strength with time (muscle strength F _{1 to} F ₆ ) of each skeletal muscle. Muscle activity is the rate at which the motor units that make up skeletal muscle are active. Therefore, the muscle activity takes a value from 0 to 1. Between the muscle activity level and the muscle strength, for example, there is a Hatze's muscle contraction model expressed by Equation (1). For example, the muscle activity calculation unit 14 calculates the muscle activity α _{1 to} α ₆ of each skeletal muscle based on the formula (1). The muscle activity levels α _{1 to} α ₆ are parameters that change over time according to changes in the coordinate position (X, Y) of the hand point 101.
F _i = T _max — _i · (α _i · k (x) · h (x ′) + f _p (x)) (1)
_{However, T Max_i} the i-th skeletal muscle maximum strength, i.e., the maximum value of strength _{F i,} k (x), by _{h (x '), f p} (x) muscle length x and velocity x' It is a determined muscle characteristic.

Here, it is known that muscle strength has a property that the degree of fluctuation in output increases as the output increases (signal strength dependent noise). Based on this knowledge, Harris et al. Assumed that noise proportional to the magnitude of the motion command (motion command-dependent noise) is superimposed, and the motion trajectory is the end point of the reaching motion where the motion command-dependent noise is superimposed. We proposed a minimum criterion for the end point dispersion that the fluctuation (dispersion) is planned to be minimized. The motion command-dependent noise by Harris et al. Is formulated as shown in Equation (2).
σ _w ² = ku ² (2)
However, w is the noise superimposed on the motion command, σ _w ² is the variance of w, u is the motion command, and k is the noise coefficient.

The end point variation evaluation unit 15 arrives when the hand 101 performs a reaching movement in a state where the motion command-dependent noise 100 is superimposed on the change in muscle activity over time (muscle activity α _{1 to} α ₆ ) of each skeletal muscle. Evaluate the end-to-end variation. The motion command dependent noise 100 corresponds to the noise w superimposed on the motion command in Equation (2). Since the muscle activity can be regarded as a motion command, from equation (2), the variance σ _wi ² of the motion command-dependent noise w _i superimposed on the muscle activity α _i is expressed as in equation (3). .
^{_{σ wi 2 = kα i 2 ...}} (3)
That is, the motion command-dependent noise superimposed on the muscle activity α _i can be simulated by randomly generating a random variable that satisfies Equation (3). At this time, the muscle activity α _i ′ to which the motion command heterogeneous noise is added is expressed by Expression (4).
α _i ′ = α _i + w _i (4)

The end point variation evaluation unit 15 generates a random random variable satisfying the expression (3) for the muscle activity α _i , that is, the motion command dependent noise 100. Then, the end point variation evaluation unit 15 superimposes the motion command-dependent noise 100 on the muscle activity α _i and causes the hand 101 to try the reaching motion a plurality of times (about several tens to several hundreds times) to reach the reaching end point of each time. Is calculated, and an error between each reaching movement end point and the reference end point of the reaching movement is calculated, and further, an average value of the errors is calculated to evaluate variations in the reaching movement end point. The error is a deviation of the reaching motion end point from the reference end point of the reaching motion, that is, a distance between the reaching motion end point and the reaching motion reference end point.

FIG. 4 shows the end point variation of a certain reaching motion when the motion command dependent noise is superimposed. The movement trajectory of the reaching movement draws a gentle arc and is considered to be a natural movement trajectory for humans. On the other hand, FIG. 5 shows a variation in the end point of another reaching motion on which motion command-dependent noise is superimposed. The movement trajectory of the reaching movement is meandering and is considered to be an unnatural movement trajectory for humans. FIG. 4A and FIG. 5A are diagrams in which a 2-joint 6-muscle upper limb model is projected on the xy plane. Joints _{J 1} is fixed at the origin of the xy plane, arriving from the hand points 101 start position along a movement path 102 (-0.27,0.3) to the end position (0.27,0.35) exercise. The end point position corresponds to the reference end point of the reaching movement. FIG. 4B and FIG. 5B are plots of the movement trajectory of the hand 101 when the above reaching movement is tried a plurality of times using the 2-joint 6-muscle upper limb model. For convenience, the motion trajectory instructed to the 2-joint 6-muscle upper limb model is indicated by a white broken line. 4 and 5, it can be seen that when the reaching motion is performed on a motion trajectory that is considered unnatural for humans, the reaching motion end point varies greatly. When the error average value of the reaching movement end point is calculated in the examples of FIGS. 4 and 5, the error average value is calculated to be larger in the example of FIG. 5 than in the example of FIG.

Alternatively, the end point variation evaluation unit 15, instead of attempting to reaching movements, for example, by performing predetermined arithmetic processing for distributed sigma _wi ² Motor Command dependent noise w _i defined in equation (3) It is also possible to evaluate the variation of the reaching movement end point. Specifically, the end point variation evaluation unit 15 calculates the time integration value of the variance of the motion command-dependent noise at the muscle activity level for each skeletal muscle, and calculates the sum of these time integration values to obtain the reaching motion end point. Can be evaluated (see equation (5)). Since the variation evaluation of the reaching movement end point by such calculation processing is a simple evaluation method to the last, there is a possibility that it does not necessarily coincide with the variation obtained by the reaching movement trial as described above. It can be used as an alternative to evaluating the variability of the end point of reaching movement.
^{_{Σ i = 1 N ∫ 0 T}} σ wi 2 dt (= Σ i = 1 N ∫ 0 T kα i 2 dt) ... (5)
However, N is the total number of skeletal muscles. In this example, N = 6 and T is the end time of the reaching movement.

Further, the end point variation evaluating unit 15 calculates the time integral value of the dispersion of the motion command dependent noise for the muscular strength for each skeletal muscle, and evaluates the variation of the reaching end point by calculating the sum of these time integral values. Yes (see equation (6)).
Σ _{i = 1} ^N _{０ 0} ^T σ _Fi ² dt (6)
However, N is the total number of skeletal muscle, N = 6, T in this embodiment is the end time of the reaching movements, sigma _Fi ² is the variance of strength F _i which is calculated based on equation (1).

  Although the presentation of the calculation formula is omitted, the end point variation evaluation unit 15 calculates the time integral value of the variance of the motion command-dependent noise for each of the joint torque, the joint angle, and the hand position for each skeletal muscle, and these time integrals. By calculating the sum of the values, it is also possible to evaluate the variation in the reaching movement end point.

Returning to FIG. 1, the reaching motion determination unit 16 changes the reaching motion of the hand in various ways, and increases the positioning accuracy for humans based on the variation in the reaching motion end point evaluated by the end point variation evaluating unit 15 for each reaching motion. Determine the reaching movement. FIG. 6 schematically shows the reaching movement of the determination target by the reaching movement determination unit 16. Here, there are two reaching movements to be evaluated, and the movement trajectory 102in of one of the reaching movements is that when the hand of the two-joint 6-muscle upper limb model moves toward the inside of the body, and the hand point 101 is the start position. It moves from (0, 0.3) to the end point position (-0.2, 0.3). The movement trajectory 102out of the other reaching movement is the same when the hand of the two-joint 6-muscle upper limb model moves toward the outside of the body, and the hand point 101 changes from the start position (0, 0.3) to the end position ( 0.2, 0.3). That is, the movement distance of the hand point 101 is the same in the movement trajectory 102in and the movement trajectory 102out. Incidentally, the joint J ₁ is fixed at the origin of the same manner as described above xy plane.

  As shown in FIG. 6, the usage rate of the muscle is almost the same in the motion trajectory 102in and the motion trajectory 102out. Therefore, when the motion trajectory 102in and the motion trajectory 102out are evaluated from the viewpoint of the conventional muscle exerting efficiency, it is determined that there is no difference between the two.

  On the other hand, FIG. 7 shows the variation of the reaching motion end point evaluated by the end point variation evaluating unit 15 for each reaching motion shown in FIG. FIG. 7A shows variations in the reaching end point of the movement trajectory 102in, and FIG. 7B shows variations in the reaching end point of the movement trajectory 102out. As can be seen by comparing FIG. 7 (a) and FIG. 7 (b), the movement end point variation is smaller when the hand of the bi-joint 6-muscle upper limb model is moved outward than when it is moved inside the body. There is a tendency.

  FIG. 8 shows, in a box-and-whisker plot, the average error value of the end points of each reaching movement shown in FIG. 7 calculated by the end point variation evaluation unit 15. The average error value of the reaching end point of the motion trajectory 102in is about 0.008, while the average error value of the end point of the reaching motion end of the motion trajectory 102out is about 0.007. P value is less than 0.001. Accordingly, the reaching motion determination unit 16 determines from the numerical evaluation results that the latter is the reaching motion that increases the positioning accuracy for humans among the reaching motion of the motion trajectory 102in and the reaching motion of the motion trajectory 102out. .

  Next, a description will be given of an experimental result when a person is actually made to reach to verify the validity of the evaluation result of the human body motion evaluation apparatus 10. FIG. 9 schematically shows the outline of the experiment. All three subjects (22 ± 1 years old) wear eye masks, move the right hand tip from the position in front of the right shoulder to the left side, that is, the motion trajectory 102 in to move inside the body, and the right hand tip in front of the right shoulder The reaching movement was performed 10 times on the movement trajectory 102out to be moved to the right side, that is, the outside of the body. The movement trajectory 102in is a movement trajectory in which the right shoulder position 101 is fixed at the origin, and the hand point 101 of the right hand moves from the start position (0, 0.2) to the end position (−0.2, 0.2). is there. The motion trajectory 102out is a motion trajectory in which the right hand hand point 101 moves from the start position (0, 0.2) to the end position (0.3, 0.2) with the position of the right shoulder fixed at the origin. . Each of the motion trajectories had an exercise time of 0.5 seconds, and 20 exercises were performed before the experiment. For the measurement of the hand point 101, an optical tracking system OptiTrack V120 manufactured by SPICE was used.

  FIG. 10 shows variations in the reaching movement end points of the reaching movements shown in FIG. FIG. 10A shows variation in the reaching end point of the movement trajectory 102in, and FIG. 10B shows variation in the reaching end point of the movement trajectory 102out.

  FIG. 11 is a box-and-whisker plot showing the average error value at the end point of each reaching movement shown in FIG. The average error value of the reaching end point of the motion trajectory 102in is about 0.04, whereas the average error value of the end point of the reaching motion end of the motion trajectory 102out is about 0.02. P value is less than 0.001. Therefore, from the result of this numerical evaluation, it is shown that the latter of the reaching motion of the motion trajectory 102in and the reaching motion of the motion trajectory 102out is the reaching motion that actually increases the positioning accuracy for humans. This experimental result is consistent with the evaluation result of the human body motion evaluation apparatus 10, and proves that the evaluation result by the human body motion evaluation apparatus 10 is appropriate.

  As described above, according to the present embodiment, the motion of the human body can be evaluated from the viewpoint of positioning accuracy. Thereby, it is possible to find a position where a human can easily position the arrangement of various parts (for example, buttons of a power window of an automobile) in a human interface of various industrial products.

  The human body motion evaluation technique according to the present invention is useful for developing human interfaces for various industrial products because it can evaluate the motion of the human body from the viewpoint of positioning accuracy.

DESCRIPTION OF SYMBOLS 10 Human body motion evaluation apparatus 11 Joint angle calculation part 12 Joint torque calculation part 13 Muscle force calculation part 14 Muscle activity calculation part 15 End point dispersion | variation evaluation part 16 Achieving movement determination part 100 Motion command dependence noise

Claims (9)

A device that evaluates human movement using a musculoskeletal model that models the human skeleton, joints, and skeletal muscles,
A joint angle calculation unit that receives a movement trajectory of the limb when the limb of the musculoskeletal model makes a reaching movement, and calculates a change in angle of each joint over time in the musculoskeletal model;
A joint torque calculation unit for calculating a temporal torque change of each joint from an angular change of each joint over time;
A muscular strength calculation unit that calculates chronological strength changes of each skeletal muscle in the musculoskeletal model from torque changes of each joint over time;
A muscle activity calculation unit that calculates a change in muscle activity over time of each skeletal muscle from a change in muscle strength over time of each skeletal muscle;
An end point variation evaluation unit that evaluates a variation in the reaching movement end point when the limb performs the reaching movement in a state where movement command-dependent noise is superimposed on a change in muscle activity over time of each skeletal muscle;
A reaching motion determination unit that changes the reaching motion of the limbs variously and determines a reaching motion that increases positioning accuracy for a human based on the variation of the reaching motion end point evaluated by the end point variation evaluation unit for each reaching motion; A human body motion evaluation apparatus comprising: The human body motion evaluation apparatus according to claim 1,
The end point variation evaluation unit
The movement command-dependent noise is superimposed on the change in muscle activity over time of each skeletal muscle and the limb is tried a plurality of times to calculate the reaching movement end point of each time,
Calculate the error between the reaching end point of each time and the reference end point of the reaching motion,
A human body motion evaluation apparatus characterized in that the average value of the errors is calculated to evaluate variations in the reaching motion end point. The human body motion evaluation apparatus according to claim 1,
The end point variation evaluation unit calculates a time integral value of variance defined for the motion command-dependent noise for each skeletal muscle in the musculoskeletal model, and calculates the sum of these time integral values to calculate the end point of the reaching motion end point. A human body motion evaluation apparatus characterized by evaluating variation. A method for evaluating human movement using a musculoskeletal model that models the human skeleton, joints, and skeletal muscles,
Receiving the movement trajectory of the limb when the limb of the musculoskeletal model makes a reaching movement, and calculating the angular change of each joint over time in the musculoskeletal model,
Calculate the torque change over time of each joint from the angle change of each joint,
Calculate the muscular strength change of each skeletal muscle over time in the musculoskeletal model from the torque change of each joint over time,
Calculate the change in muscle activity over time of each skeletal muscle from the change in muscle strength over time of each skeletal muscle,
Evaluating the variation of the reaching movement end point when the limb performs the reaching movement in a state where movement command dependent noise is superimposed on the muscle activity change of each skeletal muscle over time,
A human body motion evaluation method comprising: changing the reaching motion of the limbs in various ways, and determining a reaching motion that increases positioning accuracy for a human based on variations in the evaluated reaching motion end point for each reaching motion. The human movement evaluation method according to claim 4,
The movement command-dependent noise is superimposed on the change in muscle activity over time of each skeletal muscle and the limb is tried a plurality of times to calculate the reaching movement end point of each time,
Calculate the error between the reaching end point of each time and the reference end point of the reaching motion,
A human body motion evaluation method characterized in that a variation in the reaching motion end point is evaluated by calculating an average value of the errors. The human movement evaluation method according to claim 4,
A dispersion time integral value defined with respect to the motion command dependent noise is calculated for each skeletal muscle in the musculoskeletal model, and a variation of the reaching movement end point is evaluated by calculating a sum of these time integration values. A human body movement evaluation method. A program that causes a computer to evaluate human motion using a musculoskeletal model that models the human skeleton, joints, and skeletal muscles,
A joint angle calculation means for receiving a movement trajectory of the limb when the limb of the musculoskeletal model makes a reaching movement and calculating a change in angle of each joint over time in the musculoskeletal model;
A joint torque calculating means for calculating a temporal torque change of each joint from a temporal angle change of each joint;
Muscle strength calculating means for calculating a change in strength of each skeletal muscle over time in the musculoskeletal model from a change in torque over time of each joint;
Muscle activity calculation means for calculating a change in muscle activity over time of each skeletal muscle from a change in muscle strength over time of each skeletal muscle;
End point variation evaluation means for evaluating variation in the reaching movement end point when the limb performs the reaching movement in a state where movement command-dependent noise is superimposed on a change in muscle activity over time of each skeletal muscle, and As a reaching motion determination means for determining a reaching motion that increases the positioning accuracy for humans based on variations in the reaching motion end point evaluated by the end point variation evaluating means for each reaching motion by changing the reaching motion in various ways. A human movement evaluation program characterized by functioning. In the human movement evaluation program according to claim 7,
The end point variation evaluation means is
The movement command-dependent noise is superimposed on the change in muscle activity over time of each skeletal muscle and the limb is tried a plurality of times to calculate the reaching movement end point of each time,
Calculate the error between the reaching end point of each time and the reference end point of the reaching motion,
A human body motion evaluation program characterized by calculating an average value of the errors to evaluate a variation in the reaching motion end point. In the human movement evaluation program according to claim 7,
The end point variation evaluating means calculates a time integral value of variance defined for the motion command-dependent noise for each skeletal muscle in the musculoskeletal model, and calculates the sum of these time integral values to calculate the end point of the reaching motion end point. A human movement evaluation program characterized by evaluating variation.