A novel RBF neural network–based sliding mode controller for a master–slave motor coordinated drive system

Master–slave coordinated systems are used in various domains, but their control remains challenging. In our study, a master–slave motor coordinated drive mechanism is engineered to enhance the functionality of robotic fingers. Addressing the coordination issues between the master and slave motors, a novel RBF neural network-based sliding mode controller (RBF-SMC) is introduced. This RBF-utilized controller precisely approximates model uncertainties and external disturbances within the master–slave motor coordinated system. The integration of a hyperbolic tangent function as a substitute for the traditional switching function (RBF-SMC-Tanh) reduces chattering in sliding mode control, which is an advancement in control strategy of this motor system. Theoretical validation of system stability under the proposed controller is confirmed through the selection of an appropriate Lyapunov functional. Furthermore, simulations and experiments were carried out to compare the performance of the SMC, RBF-SMC, and RBF-SMC-Tanh controllers. The findings demonstrate that under the control of the RBF-SMC-Tanh controller, the master motor maintains accurate trajectory tracking even in the presence of force inconsistency and external disturbances. Concurrently, the slave motor adeptly follows the movements of the master motor, adhering to the predefined functional relationship of the systemic motion. The RBF-SMC-Tanh controller significantly outperforms the SMC and RBF-SMC controllers in terms of robustness, trajectory tracking accuracy, and chattering suppression. In particular, when tracking high-frequency changing target trajectories, it achieves performance improvements of 23.85–94.97% over the SMC controller and 8.64–56.28% over the RBF-SMC controller.