Weighted Multiple Model Adaptive Dynamic Surface Control for a Flexible-Joint Manipulator

Operation performance of a flexible joint manipulator is measured by the accurate tracking of angular position, which is especially significant for the plant with largely uncertain parameters. In this paper, a weighted multiple model adaptive dynamic surface control (WMMADSC) method is proposed for a flexible-joint manipulator with largely jumping parameters. The present WMMADSC approximates the largely uncertain parameters with a finite local constant parameter models. And, a local dynamic surface controller is designed to deal with the angle tracking problem, based on the local flexible joint manipulator model with constant parameters. Furthermore, a weighting algorithm is designed for fusing the local controllers to generate global control law. Simulation example and comparisons are shown to illustrate the effectiveness of the proposed WMMADSC method.