A Control Method of AC Induction Motor Based on CMAC and Single Neuron PID Controller

Abstract

A composite controller using Cerebellar Model Articulation Controller (CMAC)Neural Network and single neuron PID is proposed in the paper to solve the speed control problems between static and dynamic performance and contradiction between tracking settings and suppressing disturbance capacities due to poor robustness of conventional PID controller since AC induction motor is a high-order, strong couple, nonlinear multivariable system. The approach adopts single neuron PID controller in place of conventional PID single neuron controller parameter optimization and application of neural network controller facilitates feed forward control. On this basis, a PWM algorithm based on MCF5235 is presented, fulfilling AC induction motor governing. To assess the effectiveness of the proposed composite controller, experiments and simulations were conducted. Obtained results show the proposed approach is satisfactory in fast response, small overshoot, high control accuracy, strong anti-interference ability and robustness when compared with conventional PID controller[2].