Control of a time-delayed 5 degrees of freedom arm model for use in upper extremity functional electrical stimulation

The goal of this work is to design a controller for a functional electrical stimulation (FES) neuroprosthesis aimed at restoring shoulder and elbow function in individuals who have suffered a high-level cervical (C3-C4) spinal cord injury (SCI). The controller is a mathematical algorithm that coordinates the electrical stimulations applied to the paralyzed muscles such that the arm closely tracks a given desired trajectory. An issue that so far has received little attention is that of time-delays. These delays arise from two sources: (1) the muscle excitation-activation dynamics (10-30ms) and (2) the sampling of the electrical stimulation (80ms at the typical 12Hz stimulation frequency). Using a 5 degrees of freedom (5DOF) arm model we designed and evaluated a novel controller capable of maintaining stable and accurate tracking performance in the presence of time-delays. For a desired trajectory consisting of 10 randomized reaches, the controller achieved excellent tracking performance as measured by the root-mean-square error (RMSE) between the desired and simulated joint angles (RMSE= [1.48°; 0.81°; 2.14°; 3.11°; 2.29°]).