Analysis of Position Servo System of Pneumatic Manipulator Based on RBF Neural Network PID Control

This paper analyzes the characteristics of pneumatic position servo system of a mechanical hand in particularly respect to the nonlinearity of the position servo system of a pneumatic manipulator with 3 degrees of freedom. A pneumatic position servo model was developed in AMESim and imported into Simulink in the form of a S-function, resulting in a RBF neural network PID control system model in Simulink. Co-simulations were performed with both AMESim and Matlab/Simulink. As compared to the simulation results of the same system with AMESim model without correction, RBF neural network PID controller significantly improves the dynamic performance of the pneumatic servo system.