신장반사 해석 모델 및 진자 실험을 이용한 하지의 질량 관성모멘트 예측 방법

Abstract

It is important to predict mass moment of inertia of each segment for analyzing human motions. Analytical and experimental methods have been studied for predicting it. The former is methods calculating it by assuming the segment as uniform bar or using 3 dimensional images such as MRI, and the latter is a method predicting it from cadaver experiments. However, these methods oversimplify human segments or cost a lot of money and time. Moreover, the method using 3 dimensional images have been pointed out safety because subjects have to be exposed to radiation. Therefore, present study suggests safe and simple method that can reflect physical quality in vivo. We did pendulum tests dropping a leg relaxing and stretching in a sitting position without notice, and predict mass moment of inertia of the leg using the experimental results and stretch reflex model. The stretch reflex model consists of muscle properties and model parameters, and it can identify muscle activation by a stretch reflex from kinematics of a leg. The present studies set mass moment of inertia and model parameters as variables, and solve the optimization problem which minimizes the integrated error of knee angle from experiment and analysis to predict mass moment of inertia of a leg. Compared to results from the present study and previous study, we confirmed that suggested method is most efficient to predict the mass moment of inertia of human limb.