Repetitive Learning Control For Global Asymptotic Tracking Of Robot Manipulators

Robotic manipulator is a very complicated nonlinear system with strong-coupling characteristics. High-accuracy trajectory tracking control of robotic manipulators is especially challenging and has been widely studied in the past years. Repetitive learning control is developed based on internal model principle and accomplishes the trajectory tracking of periodic reference signal with a high precision. This approach does not utilize the modeling information in the control law formulation, and thus permits easy implementation. Given the repetitive tasks of these myriads of industrial robots, repetitive learning control has become one of the most effective solutions to high-accuracy trajectory tracking control of robotic manipulators. In this thesis, two new classes of repetitive learning controllers are proposed. The main research works can be described as follows.1. A new class of nonlinear decentralized sliding-mode repetitive learning control for robot manipulators is proposed. The control strategy is formulated with a calss of nonlinear saturated function with adjustable parameters and obtains faster response and higher tracking precision over the commonly used repetitive learning control. The favorable is achieved without increased torque. Decentralized control structure and sliding-mode control improve the fault toleration and robustness against uncertainty.2. A fast terminal sliding mode based repetitive learning control method for robot manipulators is developed by incorporating characteristics of fast terminal sliding mode control into repetitive learning control. Advantages of the proposed control include the absence of model parameter in the control law formulation and improved robustness and tracking performance in comparison with the conventional approaches.3. Lyapunov’s direct method and La Salle’s invariance principle are employed to prove global asymptotic tracking. Simulation results on a 3DOF revolute robotic manipulators demonstrate the effectiveness and improved performances of the proposed schemes.