APPLICATION OF OPTIMIZATION PRINCIPLES IN DETERMINING APPLIED MOMENTS IN HUMAN LEG JOINTS DURING GAIT

A method is developed based on the principles of optimization to obtain the approximate joint moments during gait directly from a set of measured displacement history. This type of problem is defined as the inverse dynamics problem. Compared with the conventional techniques currently used by many investigators in solving inverse dynamics problems, the present method is more advantageous because it requires no numerical differentiation of experimental data. Thus, the possibility of magnifying inherent experimental errors is minimized. Starting with reasonable initial estimates, it is possible to obtain an approximate solution which is within 5% of the exact solution everywhere throughout the entire time interval investigated. If the set of applied moments in a given gait problem is expected to be unique corresponding to the set of measured displacements, then the proposed method will improve the initial estimates significantly and yield final solutions which are fairly close to the correct solution, even when the initial estimates differ considerably from the correct moments.