A Neuromuscular-Model Based Control Strategy to Minimize Muscle Effort in Assistive Exoskeletons

In literature, much attention has been devoted to the design of control strategies of exoskeletons for assistive purposes. While several control schemes were presented, their performance still has limitations in minimizing muscle effort. According to this principle, we propose a novel approach to solve the problem of generating an assistive torque that minimizes muscle activation under stability guarantees. First, we perform a linear observability and controllability analysis of the human neuromuscular dynamic system. Based on the states that can be regulated with the available measurements and taking advantage of knowledge of the muscle model, we then solve an LQR problem in which a weighted sum of muscle activation and actuation torque is minimized to systematically synthesize a controller for an assistive exoskeleton. We evaluate the performance of the developed controller with a realistic non-linear human neuromusculoskeletal model. Simulation results show better performance in comparison with a well known controller in the literature, in the sense of closed loop system stability and regulation to zero of muscle effort.