Composing an Assistive Control Strategy Based on Linear Bellman Combination from Estimated User's Motor Goal

Abstract

In assistive control strategies, we must estimate the user's movement intentions. In previous studies, such intended motions were inferred by linearly converting muscle activities to the joint torques of an assistive robot or classifying muscle activities to identify the most likely movement from pre-designed robot motion classes. However, the assistive performances of these approaches are limited in terms of accuracy and flexibility. In this study, we propose an optimal assistive control strategy that uses estimated user movement intentions as the terminal cost function not only for generating movements for different task goals but to precisely enhance the motion with an exoskeleton robot. The optimal assistive policy is derived by blending the pre-computed optimal control laws based on the linear Bellman combination method. Coefficients that determine how to blend the control laws are derived based on low-dimensional feature values that represent the user's movement intention. To validate our proposed method, we conducted an assisted basketball-throwing task and showed that the performances of our subjects significantly improved.