Research on Ant Colony Neural Network PID Controller and Application

NN (neural network) combine with traditional PID (proportional integral derivative) control to make control system has corresponding degree aptitude. However, NN rate of convergence is slower and is liable to get into local minima and affect NN application in the real time control. In order to search speediness algorithm of global convergence to satisfy the real time control and better performance, this paper applies ACA (ant colony algorithm) to optimize the parameters of NN-PID controller to improve the on-line self-tuning capability of this controller. At the same time, the strategy is implemented using TMS320F240 digital signal processor on induction motor drive DTC (direct torque control) system. Experiment results validate that this method is validity and the system has a better dynamic and static state performance.