Abstract：This paper presents a class of discrete-time nonlinear systems model-free adaptive sliding mode control algorithm to solve the mathematical model dependency and model-free dynamics of the controlled system, and improved its control law. This control strategy belongs to a kind of data-driven control method in essence, and which controller designing depends on the system input and output data without any controlled system model information. Theoretical analysis showed that this proposed control algorithm enhanced the utilization of pseudo partial derivative estimate information and accelerated the convergence's velocity, and the stability of control system was verified by strict mathematical deduction. This proposed model-free adaptive sliding mode control algorithm has obtained better control performance through the simulations for linear motor position and speed information tracking, which has achieved accurate traceability and strong robustness.

Keyword：nonlinear system;model-free adaptive control;sliding mode control;pseudo partial derivative;