International Journal of Control, Automation and Systems October 2017,
Abstract: In this paper, a two-wheel robot (TWR) system is implemented based on a position and angle decoupled intelligent backstepping control system. The decoupled method provides a simple method to achieve asymptotic stability control for a fourth-order nonlinear TWR system using a second-order dynamics model. Moreover, adaptive output recurrent cerebellar model articulation control (AORCMAC) is designed in this study to imitate an ideal backstepping controller (IBC). Furthermore, to compensate for the difference between the two controllers mentioned, a compensator is used. Based on the Lyapunov stability method, the stability of the closed-loop system can be guaranteed. To train the AORCMAC effectively, the variable learning rates that guarantee the convergence of the system are proposed. Finally, simulation and experimental results demonstrate the effectiveness of the proposed control method for TWR control.