A servo system control research based on fuzzy CMAC

The conventional PID controller for servo system can hardly restrain torque disturbance caused by friction and load efficiently, and it has slow respond speed. This paper introduces a compound control method for servo system based on fuzzy CMAC neural network. FCMAC neural network introduce membership function and activation intensity of memory cells, which makes the description ability of network stronger and the approach precision of network will be better. The controller based on FCMAC neural network can apperceive disturbance by self learning on line and provide appropriate compensation for disturbance. The results of simulation show that with this method servo system will have better tracking ability and well robustness.