A Method Of Optimization Based Human Dynamic Simulation For Exoskeleton Robot Design And Assessment

In this paper we formulate optimization based mathematical model of human dynamic motion prediction which can be used as an exoskeleton robot design and assessment tool. With this approach, motion is governed by human performance measures which is act as objective functions to be minimized in the formulation. As an exoskeleton robot model, external forces to the skeletal model of human body are applied. Lifting task is adopted and formulated as a numerical example. The human model has 55 degrees of freedom - 49 revolute joints and 6 global translation & rotation joints. An algorithm based on the sequential quadratic programming approach is used to solve the nonlinear optimization problem, and analytical gradients for the optimization process are calculated and provided. The results with joint torques and mechanical energy of human over the motion are recovered and analyzed under various cases. It shows that the effect of exoskeleton robot to the human body and proves the ability of this research as an exoskeleton design and assessment tool.