Switched control of an N-degree-of-freedom input delayed wearable robotic system

Abstract

In this paper, a switched control method for a class of wearable robotic systems that prioritizes the use of human skeletal muscles in an assistive rigid powered exoskeleton is derived. A general N-degree-of-freedom (N-DOF) human–robot model is proposed to consider the challenges induced by the wearable system that include uncertainties and nonlinearities, unilateral actuation properties of the skeletal muscles, input delays, as well as a time varying actuator efficiency. Two control modes that alternatively switch and control a wearable robotic system are designed to overcome these challenges. A multiple Lyapunov functional analysis with state-dependent constraints on the switch criteria is performed to prove the stability. Simulations are performed to demonstrate the gain conditions, selected for each subsystem, that stabilize the overall system. Experiments on a human participant wearing a 4-DOF hybrid exoskeleton that combines functional electrical stimulation and a powered exoskeleton demonstrate the effectiveness of the switched control design.