A Human Motor Control-Inspired Control System for a Walking Hybrid Neuroprosthesis

摘要

The purpose of this research is to develop a human motor control-inspired control system for a hybrid neuroprosthesis that combines functional electrical stimulation (FES) with electric motors. This device is intended to reproduce gait for persons with spinal cord injuries (SCI). Each year approximately 17,000 people suffer from an SCI in the U.S. alone, of which about 20% of them are diagnosed with complete paraplegia. Currently, there is a lot of interest in gait restoration for subjects with paraplegia but the existing technologies use either solely FES or electric motors. These two sources of actuation both have their own limitation when used alone. Recently, there have been efforts to provide a combination of the two means of actuation, FES and motors, into gait restoration devices called hybrid neuroprostheses.;In this dissertation the derivation and experimental demonstration of control systems for the hybrid neuroprosthesis are presented. Particularly, the dissertation addresses technical challenges associated with the real-time control of a FES such as nonlinear muscle dynamics, actuator dynamics, muscle fatigue, and electromechanical delays (EMD). In addition, when FES is combined with electric motors in hybrid neuroprostheses, an actuator redundancy problem is introduced. To address the actuator redundancy issue, a synergy-based control framework is derived. This synergy-based framework is inspired from the concept of muscle synergies in human motor control theory. Dynamic postural synergies are developed and used in the feedforward path of the control system for the walking hybrid neuroprosthesis. To address muscle fatigue, the stimulation levels are gradually increased based on a model-based fatigue estimate. A dynamic surface control technique, modified with a delay compensation term, is used to address the actuator dynamics and EMD in the control derivation. A Lyapunov-based stability approach is used to derive the controllers and guarantee their stability. The outcome of this research is the development of a human motor control-inspired control framework for the hybrid neuroprosthesis where both FES and electric motors can be simultaneously coordinated to reproduce gait. Multiple experiments were conducted on both able-bodied subjects and persons with SCI to validate the derived controllers.