An RBFNN-Informed Adaptive Sliding Mode Control for Wheeled Mobile Robots

This paper deals with the trajectory tracking of differential driven wheeled mobile robots (WMRs). It is widely known that parameter uncertainty and unknown external disturbances could bring adverse effects to WMRs. To achieve high-performance tracking, a control scheme is proposed in this paper, which integrates a kinematics controller and a dynamics controller. The kinematics controller is designed by the backstepping control and the dynamics controller is derived via adaptive sliding mode control. Moreover, dual radial basis function neural networks (RBFNNs) with adaptive adjust algorithms are adopted to approximate parameter alterations of system dynamics as well as external disturbances. To solve the uncertainty of wheel rolling radius, an adaptive control law is applied. In addition, a PI-type sliding mode control strategy with an exponential reaching law is designed to compensate for the approximation errors of the RBFNNs. Furthermore, the stability and global asymptotic convergence of the control system are proved by the Lyapunov stability theorem. Finally, simulations are provided to verify the effectiveness of the proposed control strategy. The results indicate that the proposed control law is effective in response to parameter uncertainty and external disturbances.