PRINCIPLES OF ADEQUACY CRITERIA FORMULATION IN HUMAN MOTION ANALYSIS

Abstract

Introduction. Number of parameters of an anthropomorphic model (AM), which simulates real human motion, can achieve the value of one hundred and even more than that. This makes obvious the necessity of adequacy criteria formulation. Optimal value of such criteria should indicate structural and parametric adjustment of AM to certain real human motion. Modelling of human motion with employment of mechanical-mathematical apparatus of system of body motion equations implies a significant number of problem parameters [1] required for description of the structure, and components and kinematics of motion as well. Choice of these parameters values seriously depends on what experimental data is available. METHODS AND RESULTS: The base of computer model consists in a system of differential-algebraic equations of motion of a ramified kinematics chain with nonstationary constraints. In particular, as constraint equations there can serve generalized coordinates behaviour functions, obtained through video-registration data processing. Such approach allows to determine main dynamic values, including generalized forces. However, measurement errors lead to significant errors in assessed values of inter-element forces and moments and especially values of external with respect to AM ground reaction and total moment of external forces in support phase of motion. Variation of AM elements parameters, positions of joints, parameters of trajectories smoothing allows to obtain an averaged assessment of external forces values. In the report there is suggested a new approach to structural an parametrical adjustment of AM. Presence of non-stationary constraint equations allows to use some of experimental data for such constraints. For example, ground reaction force and/or external moment can be available or equal to zero during the flight phase. One of investigation result is that there have been analyzed grand circles on the horizontal bar with a following jump off the bar and four backward somersaults performed in a grouped position. The number of AM elements is widely varied. There has been investigated influence of possible errors in determination of visco-elastic properties of the bar on the analysis results for different processing procedures. CONCLUSION: The suggested approach to iterational parametric adjustment of AM on the basis of employing of constraint equations allows for complete matching of model motion characteristics with most important experimental data. Less important data are estimated in average, which corresponds to traditional structural- parametric adjustment of AM. REFERENCES: 1. Zinkovsky A.V., Sholuha V.A., Ivanov A.A. Mathematical Modelling and Computer Simulation of Biomechanical Systems, WSP, Singapore, 1997. 216p.