Neural Network Sliding Mode Control for Pneumatic Servo System Based on Particle Swarm Optimization

Abstract

Problems of pneumatic servo system, such as poor stability, non-linearity and uncertainty in the process of modeling, seriously affect the development of high-performance pneumatic controller. In this paper, a pneumatic servo system’s mathematical model is established at first and locally linearized to a third-order nonlinear system to simplify it. Then, to eliminate the chattering problem of sliding mode control, RBF neural network is applied to approximate the control law. Besides, particle swarm optimization (PSO) is come up to achieve the overall optimization effect of RBF neural network and further improve the control performance. Lyapunov function is defined to verify the system’s stability. The results of simulation show that the neural network sliding mode controller optimized by PSO overcomes the chattering problem of pneumatic actuator, ensuring the stability, robustness and rapidity of the pneumatic system.