Auto-Tuning Fuzzy Force/Position Control Of A 5 Dof Exoskeleton For Upper Limb Rehabilitation

Active exoskeleton controllers are designed to control the interaction force. Compliant exoskeletons need to be controlled in both position and force due to the coupling with a human in all rehabilitation conditions, which give rise to human-exoskeleton interaction dynamics, high nonlinear uncertain exoskeleton dynamics, noisy sensors and other parametric uncertainties, such as environmental contacts. These factors do not allow to account on a precise dynamical model, thus model-based controllers are difficult to implement. This paper presents an auto-tuning force/position controller based on fuzzy inference system to control a previously designed five degrees of freedom upper limb exoskeleton. The exoskeleton control system aims to achieve the best control performances together with stability and safety guarantees.