The Application of Rotor Position Sensorless-Detection of Switched Reluctance Motor by RBF Neural Network Based On PSO Algorithm

Due to the rotor position of switched reluctance motor (SRM) was a highly nonlinear function of stator windings current and flux linkage, so general linear and analytical methods were difficult to achieve precision results, a method was presented that a RBF neural network based on Particle Swarm Optimization (PSO) algorithm was used to rotor position sensorless detection of SRM. The structure and parameters of RBF neural network were optimized by subtractive clustering algorithm and PSO algorithm, hence, extensive mapping ability of neural network and rapid global (local) convergence of PSO algorithm were fully developed. The simulink was carried out based on the Matlab7.1.The neural network model was simulated for finding the rotor position at different currents from a suitable measured data for a given SRM. In order to testify the validity and accuracy of the model, the on-line experiment was carried out. Results of experiment show that the scheme is simple, it has high precision, and it has fast convergence and can accurately detect the rotor position.