Multiple-Model Adaptive Control of Functional Electrical Stimulation

Abstract

This paper establishes the feasibility of multiple-model switched adaptive control to regulate functional electrical stimulation for upper limb stroke rehabilitation. An estimation-based multiple-model switched adaptive control (EMMSAC) framework for nonlinear time-invariant systems is described, and extensions are presented to enable application to time-varying Hammerstein structures that can accurately represent the stimulated arm. A principled design procedure is then developed to construct both a suitable set of candidate models from experimental data and a corresponding set of tracking controllers. The procedure is applied to a sample of able-bodied young participants to produce a general EMMSAC controller. This is then applied to a different sample of the population during an isometric nonvoluntary trajectory tracking task. The results show that it is possible to eliminate model identification while employing closed-loop controllers that maintain high performance in the presence of rapidly changing system dynamics. This paper hence addresses critical limitations to effective stroke rehabilitation in a clinical setting.