Multiple model based adaptive control of robotic manipulators

A novel adaptive control strategy is proposed to improve the overall tracking performance of robotic manipulators with unknown or changing dynamics. The proposed method makes use of multiple models of the manipulator for identification. The models used are identical, except for the initial estimates of the unknown inertial parameters of the manipulator and its load. The mathematical formulation of the controller algorithm and the stability analysis of the overall system are introduced. Simulation and experimental studies are also included to demonstrate the improvement in the tracking performance.<>