In order to overcome the drawback of PID controller,effectively curb the overshoot problems of state variables,speed up the control speed of the system and enhance the antiinterference of the system,a selflearning sliding mode control (SLSMC) method was proposed for the fly ash constant pressure conveying system.A nonlinear smoothing function (NSF) was designed and then applied to the design of active disturbance rejection controller,tracking differential and sliding mode reaching law.To further improve the adaptive control ability of the system,the steepest descent method was used to update the gain parameters of the sliding mode controller through selflearning algorithm.Simulation and experimental results show that the proposed method not only has fast response speed and high control precision,but also effectively suppresses and eliminates the system chattering and the overshoot phenomenon.