Rotor position estimation of high-speed SRM drive using neural networks

Based on the artificial neural networks(ANN), a new rotor position estimation method for switched reluctance motor(SRM) drives is investigated in this paper. The nonlinear magnetic characteristics of SRM, obtained by finite element analysis(FEA), are used as the training data. After sufficient training, the correlation among flux linkage, phase current and rotor position can be built up with the ANN. In this way, the rotor position of SRM can be estimated with the ANN from the calculated flux linkage and phase current. In order to verify the validity of this method, a 7.5 kw 12/8 structure high-speed SRM is designed as the study object and the corresponding simulation model is developed via Matlab. Simulation results testify the validity of this algorithm.