Real-Time Learning Controller Design for a Two-Link Robotic Arm

Abstract

In this paper, a real-time learning control method involving the proportional-derivative controller and cerebellar model articulation controller (CMAC) is proposed. A feed-forward compensation using CMAC is proposed to learn and control the uncertain system dynamics with unknown but bounded nonlinearities. A priori knowledge of the system parameter values is not required. An application of robotic arm control system is carried out to demonstrate the effectiveness and robustness of the control method.