SRM Modeling And Rotor Position Detection Based On Optimized LSSVM

Switched Reluctance Drive(SRD)is a type of excellent performance's speed regulation system.It consists of Switched Reluctance Motor(SRM),power converter,controller module and position detection module,has the characteristics of simple structure,reliable operation,flexible control,and is widely used.This paper summarizes the research status of flux linkage modeling and position-free detection technology for switched reluctance motor,taking three-phase(12/8 pole)SRM as the research object,the basic structure and working principle of SRM are analyzed,and the curve of SRM nonlinear flux characteristic and torque characteristic is obtained by the test measurement method,and the SRM flux characteristic model is established.On the basis of the flux linkage model,the position detection method without position sensor is studied.Because SRM operates in a state' where the magnetic circuit is highly saturated,its magnetic flux characteristics are highly nonlinear,and it is difficult to obtain accurate mathematical models.In this paper,the nonlinear mapping capability of the least squares support vector machine is used,and the magnetic flux model and torque model of the SRM are established by training and learning the sample data.In order to improve its generalization ability and prediction accuracy,the particle swarm optimization algorithm was used to optimize the LSSVM.Through simulation verification,the accuracy was significantly improved before and after optimization.Then,based on the optimized least squares support vector machine,a position estimation model structure with phase winding flux and current as input and rotor position as output is established.The estimation model realizes the indirect detection of the rotor position of the SRM through MATLAB/Simulink simulation platform,and the optimization of the particle swarm algorithm improves the accuracy of the rotor position estimation.Finally,the SRM driving system experiment platform is built to verify the scheme.The experimental results show that the optimized estimation model of LSSVM realizes the detection of rotor position,improves the accuracy compared with that before optimization.