Modeling and simulation of switched reluctance motor based on RBFN-AFS

Considering the nonlinear, saturation and coupled magnetization, this paper presents a radial basis function network-based adaptive fuzzy system (RBFN-AFS) to model the switched reluctance motor (SRM) and predict the performance in SRM drive system. Based on the measured SRM's flux linkage and torque data, the RBFN-AFS is designed to learn and train the electromagnetic characteristics knowledge for the SRM by using the hierarchically self-organizing learning (HSOL) algorithm to determine the minimum necessary number of rules and adjust the mean and variance vectors of individual hidden nodes as well as their weights. After training, the RBFN-AFS forms a very efficient mapping structure for the nonlinear characteristics of the SRM. Lastly, a RBFN-AFS current-dependent inverse flux linkage model and a RBFN-AFS torque model are used to simulate the dynamic performance of a 6/4 0.55 kW SRM. The simulation results and experimental waveforms are reported to validate the proposed RBFN-AFS modeling method for SRM. It also provides the application of analysis and real time control for SRM.