Adaptive Backstepping Control for Mecanum‐Wheeled Omnidirectional Vehicle Using Neural Networks

In this paper, an adaptive backstepping controller is designed for the mecanum‐wheeled omnidirectional vehicle (MWOV) with frictions, external disturbances, and unknown slipping parameters. First, kinematic and dynamic models of the MWOV with wheels' slipping are derived. Especially, the virtual neural networks are introduced to on‐line approximate the nonlinear uncertainties of the dynamic system. Then, the adaptation laws with σ ‐modification are employed to estimate the unknown parameters. Furthermore, based on the backstepping technology, an adaptive backstepping controller is synthesized with the updated values, and stability and robustness of the proposed control scheme are analyzed via the Lyapunov theory. Finally, comparative simulation results demonstrate the effectiveness of the proposed method.