Abstract: In this paper, the Lyapunov function of the nonlinear sliding mode is trained by the learning algorithm of the perceptron which has the characteristics of the linear separable classification. The switching function of the nonlinear system is obtained by this Lyapunov function and the sliding mode controller is designed also by the perceptron constructed with this switching matrix. A new systematic procedures of the nonlinear system is developed by this perceptron based nonlinear sliding mode control. The simulation of the turning system using this control strategy shows that this control strategy is much efficiency.