Nonlinear Modeling And Torque Ripple Suppression Of Switched Reluctance Motor Used In BSG System

In order to improve the shortcomings of high carbon emissions of traditional oil vehicles,and short endurance of pure electric vehicles,hybrid vehicles,which take into account the two advantages of low carbon and long cruising range,have received widespread attention.The belt-driven starter generator(BSG)is one of the most popular technology in the micro-hybrid system.Among the available BSG motors,SRM has better development prospects due to its advantages of no permanent magnets,simple structure,and large starting torque.This thesis was researched under the funding of the National Natural Science Foundation of China Project 51875261 and the Jiangsu Outstanding Youth Fund Project BK20180046.In order to improve the torque calculation accuracy and ripple of a four-phase 16/10-pole segmented-rotor switched reluctance motor(SSRM)used in the BSG system of hybrid electric vehicles,researches were carried out in this paper,including nonlinear modeling,direct torque control,sliding mode control,simulation analysis based on Simulink,and experimental verification of dSPACE rapid control prototype platform.The main researches were as follow:1.Based on the introducation of structure of the four-phase 16/10-pole SSRM,the operating principle of SSRM is explained.According to the simplified two-port device model,the electromechanical energy conversion principle of SSRM was revealed from the voltage equation,mechanical motion equation and electromechanical integrate equation.Described the establishment methods of three mathematical models of SSRM,and explain why choosed the nonlinear model.Finally,the basic speed control mode of SSRM was explained,which provided a theoretical basis for the subsequent simulation and verification research of nonlinear models and control methods.2.To establish the nonlinear model of SSRM,the flux linkage-current-angle and torque-current-angle data of SSRM were obtained through finite element analysis.Maximum-correntropy-criterion-based least squares support vector regression(MCC-LSSVR)was introduced to regress the nonlinear relationship between flux linkage-current-angle and torque-current-angle.Then,a torque model that can be used to quickly and accurately calculate torque was formed.The proposed MCC-LSSVR has adaptive weights,which can reduce the interference of outliers.The gray wolf algorithm is used to optimize the hyperparameters in MCC-LSSVR to reduce the calculation time.Finally,Simulink was used to simulate and verify the proposed nonlinear model.3.Aiming at the problem of large torque ripple of SSRM,combined with sliding mode controller and observer,an improved direct torque control(DTC)was proposed to suppress torque ripple.Aiming at the modest performance of traditional PI speed controllers in anti-interference and nonlinear control in DTC systems,a sliding mode speed controller based on a new reaching law was proposed.An anti-disturbance sliding mode observer was proposed and combined with the sliding mode speed controller to establish an anti-disturbance compound speed control strategy.The real-time torque in DTC was calculated by the MCC-LSSVR torque model.Finally,Simulink was used to verify the effectiveness of the proposed new reaching law,sliding mode speed controller and anti-disturbance sliding mode observer.4.Designed a rapid control prototype development platform based on dSPACE.Introduced the process of the platform and the corresponding hardware design.Specifically include: sensor signal acquisition and conditioning circuit,protection circuit,isolation circuit,power switch drive circuit and voltage conversion circuit,etc.Finally,built a rapid control prototype development platform based on dSPACE to prepare for SSRM experimental research.5.Using the designed rapid control prototype development platform based on dSPACE,an SSRM experimental platform was built,and experiments for verifying the nonlinear model based on MCC-LSSVR and the proposed improved DTC were carried out.The experimental results matched well with the simulation results,and the effectiveness of the theory was verified.