Hybrid robust tracking control for a mobile manipulator via sliding-mode neural network

This paper develops a methodology for trajectory tracking control of a nonholonomic wheeled mobile manipulator with uncertainties and external load changes. The proposed control law consists of two levels: kinematics and dynamic levels. First, the auxiliary kinematic velocity control laws for the mobile platform and the onboard manipulator are separately proposed via backstepping. Then, a hybrid robust tracking control system is presented to ensure the velocity tracking ability in spite of the uncertainties. To achieve the goal, a neural network controller is developed to mimic an equivalent control law in the sliding-mode control, a robust controller is designed to incorporate the system dynamics into the sliding surface for guaranteeing the asymptotical stability, and the proportional controller is designed to improve the transient performance for randomly initializing neural network. All the adaptive learning algorithms for sliding-mode neural networks (SMNN) are derived from the Lyapunov stability theory so that the system tracking ability can be guaranteed in the close-loop system no matter the uncertainties occur or not. Simulation results illustrate the feasibility as well as efficacy of the proposed control strategy in comparison with the backstepping method.