An adaptive sliding mode controller for electro‐hydraulic servo systems combining offline training and online estimation

Abstract

In order to improve the tracking accuracy of electro‐hydraulic servo systems under nonlinear disturbance, an adaptive sliding mode controller (SMC) based on generalized regression neural network (GRNN) is proposed. The nonlinear factors and external disturbances of systems are considered in the controller, and an improved GRNN is used. In addition, the neural network achieves nonlinear approximation of the unknown part by online learning, determines the parameters of the SMC in real time by training the model offline, and reduces the impact of online estimation errors on the system to improve control accuracy. Finally, the effectiveness of the control method is verified by simulation.