Switched Reluctance Motor Direct Torque Control Research Based On RBF Neural Network

Switched Reluctance Motor(SRM) is a new type of electric machine that is simply structured, lowly priced and reliable operation. Due to its doubly salient structure and operating mode, there are several inevitable questions during its promotion, such as large torque pulsation and too much noise. These questions have restricted the promotion of this motor. In this paper, the overall performance of the switched reluctance motor drive is enhanced through the improvement of the control strategy.In this paper, on the basis of SRM mathematical model and in accordance with the principle and method of direct torque control which has been applied in asynchronous motor field, the switched reluctance motor is built based on DTC simulation model in the Matlab/Simulink environment. Then the direct torque control system is compared with the speed-regulating system in the traditional controlling mode. The simulation results show that the speed-regulating system under direct torque control is more superior.In order to further improve the performance of the direct torque control system of switched reluctance motor, the originally fixed parameter PID is replaced by fuzzy adaptive PID system; and the original table switch is replaced by RBF neural network state selector. The fuzzy adaptive PID improved the insufficient regulating ability of the fixed performance PID system in face of non-linear and controlled parameter changing systems; the features of RBF neural network, such as the ability of parallel computation, the stronger fault-tolerant ability and robustness, improved the stability of the system. It is suggested by the simulation results that the new speed-regulating system shows dynamic and static property that is obviously superior to the system under the traditional direct torque control. Moreover, the robustness and self-adaptive ability of the system are significantly improved.