Research On Switched Reluctance Motor Nonlinear Modeling And High-Performance System

Problems have been found in analyzing, designing and controlling the switched reluctance motor (SRM) because of its highly nonlinear magnetic characteristics. Accurate characteristics of flux-linkage and torque are very important for model and high-performance driving. Based on indirect measurement of flux-linkage, this paper presents an experimental method for accurate characteristics of SRM. Variation data of voltage and current were measured by DSP, then communicated to upper computer which received data with MSComm control in VB. Accurate characteristics of flux-linkage, torque and inductance were achieved by calculating, analyzing and drawing in MATLAB.Building accurate flux-linkage model of SRM is the key to improve performance of switched reluctance driving (SRD).After measuring the accurate flux-linkage data, the nonlinear model of SRM was developed by BP neural networks (BPNN) based on Levenberg-Marquardt arithmetic. This network model is of fast training convergence, generalization and small network scale handy for real-time control. The BPNN model is prior to several models. A complete dynamic simulation model of SRD on MATLAB was developed. This provides potential approach for research of parameters optimization and control strategy, as well as analysis of dynamic and static performance. Compared simulation current with measured motor phase current, accuracy of the dynamic simulation model was proved.To overcome the disadvantage of higher torque ripple in SRM, one method of instantaneous torque control by optimum profiling of the phase currents is presented. The torque ripple minimization can be achieved by optimum profiling of the phase currents based on the BPNN torque reverse model. Simulation results verify the feasibility of this torque ripple minimization technique.After the influence of current-reflowing means as well as turn-on angle and turn-off angle for efficiency was studied, a high performance SRD was presented to have both minimization torque ripple and maximization efficiency. The high performance system is verified feasibility by the simulation results.Because of its nonlinear with multivariable and strong coupling, both parameters and structure of SRD are variable under different control mode. So conventional PI controller cannot meet ideal dynamic performance requirement. Fuzzy controller was designed to improve this problem. In view of the problem of lower controlling precision of conventional fuzzy controller, a variable universe fuzzy controller in which universe was contracted along with error reducing was presented for high precision control.