Implementation and experimental tests of an impedance control of pneumatic artificial muscles for isokinetic rehabilitation

Abstract

The impedance control of an pneumatic artificial muscle for isokinetic rehabilitation applications was investigated. Due to the direct contact of rehabilitation robots with the human body, the safety and reliability of these robots are of significant importance. For such applications, position and force must be introduced simultaneously. In this regard, there are two scenarios in which the controller should compromise between the position commands and the resultant force or vice versa. To achieve these goals and considering the safety requirements, a novel control algorithm was proposed, which was identified as a reliable strategy that can be utilized for rehabilitation purposes. The fuzzy sliding mode controller was implemented to control the actuator’s velocity (time-varying position signal) to eliminate/reduce chattering and improve the settling time. The performance of the controller using the position-based impedance algorithm was investigated through lab experiments. For experiments, a mechanism consisting of a pneumatic muscle and a pneumatic proportional valve was used, and the control of vertical reciprocating motion of a mass attached to the end of the muscle was investigated. Three different reference signals were used in tests, and the maximum tracking error was measured to be less than 6%. Using the measured error criteria, a parametric study was performed to identify the effects of three impedance parameters on the outputs. The results of the parametric study were reported through response surfaces and sensitivity charts. The presented methods have not been implemented in prior research for controlling a pneumatic muscle, and the results of the experiments were satisfactory.