Akademik Veri Yönetim Sistemi
JOURNAL OF THE BRAZILIAN SOCIETY OF MECHANICAL SCIENCES AND ENGINEERING, cilt.45, sa.6, ss.1-17, 2023 (SCI-Expanded)
In this study, an intelligent adaptive interaction torque control for a pneumatically actuated forearm rehabilitation robot has been proposed. In this context, a cascade fuzzy adaptive controller has been designed to get stable interaction torque fields at changing levels to provide a haptic environment. To improve the efficiency of the controller, non-linear friction torque identification of the pneumatic actuator based on changing operating conditions has been conducted. Parallel to this, a user motion intention detection algorithm has been designed to provide compliant, safe and suitable human-robot interactions. The empirical corrective signals have been found and applied in the control algorithm to eliminate the disturbance cases at low interaction torque levels. Stability analysis has been performed at the boundary-input boundary-output (BIBO) stability conditions. To show the superior performance of the proposed cascade fuzzy adaptive algorithm, a PID algorithm has also been designed. A variety of experimental comparison tests involving a healthy user have been performed based on torque trajectory tracking performance in Hardware-in-the-Loop environment. The proposed control technique has better convergence dynamics up to ±2 % tracking errors when compared to cascade PID algorithm from performance metrics point of view.