Intelligent Control Of Switched Reluctance Motor Based On Fuzzy FCMAC Neural Network PID

Switched reluctance drive has many advantages such as simple structure, solid and reliable, low cost etc.Its outstanding characteristic is high efficiency, wide range of speed, energy-saving effect, great start torque, no startup current impulse, control-flexible. Therefore it received widespread concern in recent years.But switched reluctance motor is a time-varying, nonlinear, multivariable system, and the people are not able to get its accurate mathematical model yet, so the conventional linear controller is difficult to meet dynamic and static performance requirements of switched reluctance motor control system, which its control effect also has not been very good. Therefore this dissertation is focused on theory of intelligent control and control system design for switched reluctance motor.This dissertation based on work-principle and speed-characteristics of switched reluctance motor, and made a comparative study with other speed system. It made a selective analysis about the improvement performance to achieve optimal control of switched reluctance motor. Then based on the single fuzzy control and neural network control,combined local approximation, simple and quick cerebellar model neural networks with fuzzy logic.And designed a kind of fuzzy FCMAC neural network controller, Fuzzy reasoning was realized the cerebellar model neural network in this controller. So in the traditional neural network without definite physical meaning was endowed with reasoned parameters of fuzzy logic.This controller was stronger self-learning capacity and reflect continuity and vague of the brain cognitive preferably, it overcomed shortcomings of neural network control and fuzzy control.And made the contrastive simulation about the controller and general neural network controller, got relatively ideal control effect.Finally, Aimed at the traditional PID of SRM was not good enough, and puted forward switched reluctance motor intelligent control based on fuzzy FCMAC neural network PID, Real-time adjusted PID control parameters, and built switched reluctance motor control system simulation model in MATLAB/Simulink, made the simulation contrast analysis about steady and dynamic performance with traditional control methods, The simulation results showed that it compared with the traditional PID control method, the control method greatly improved dynamic and static performance of SRD, torque ripple was small,it did not require accurate mathematical model and had high control accuracy, small overshoots, high robustness to disturbances.