Research Of Iterative Extended State Observer For Switched Reluctance Motor Drive

Combining the advantages of DC motor and AC motor,switched reluctance motor(SRM)has the advantages including simple structure,low manufacturing cost,flexible control methods,high working reliability and high efficiency,which makes it widely used in electric vehicle driving,household appliances,aviation industry,servo systems and other fields.However,due to the double salient pole structure,SRM has a problem of severe torque ripple when it works,which is also the main obstacle confining the promotion of SRM.Therefore,the research on reducing the torque ripple of SRM is of great practical significance.Aiming at the high nonlinearity of the SRM model and with the help of iterative extended state observer(IESO),this paper discusses two kinds of iterative learning control(ILC)methods which are Active Disturbance Rejection based Iterative Learning Control(ADR-ILC)and simplified IESO based Data Driven Optimal Iterative Learning Control(simplified IESO-DDOILC).Compared with traditional ILC,these two control methods quickly estimate and compensate the uncertainties in control system and get faster convergence and smaller tracking error.Based on these two methods,this paper designs the torque compensator and current controller of SRM and reduce the torque ripple of SRM and improve the control accuracy through Simulink experiments.Finally,on the real-time simulation system d SPACE,a switched reluctance drive(SRD)is built for physical experiments.The development,advantages,disadvantages,and a series of related research of SRM is introduced in this paper firstly.Then,the mathematical model and related simplified models of SRM are analyzed.Three traditional control strategies of SRM are introduced.Secondly,the basic structure of IESO is given and its ability to estimate uncertainties is analyzed.Thirdly,based on IESO,two iterative learning control methods are proposed.The first one is ADR-ILC,which is designed in iteration domain to compensate the total uncertainties estimated by IESO by reference to the practice of ADRC in time domain.The second one is simplified IESO-DDOILC.Combining the advantages of model free adaptive control(MFAC),ILC and ESO,thi me h d li ea i e he li ea c lled la i a af e form related to control input and incorporate total uncertainties into a nonlinear term which is estimated by IESO.Finally,to apply these two methods to the actual application of SRM,this paper designs the torque compensator based on ADR-ILC and the current controller based on ADR-ILC and IESO-DDOILC h gh MATLAB Sim li k la f m.I i h ha he torque compensator can estimate and compensate the nonlinearity of the torque-current model effec i el a d ha e e fa e e eed.Wha m e,he IESO ba ed c e c lle achieves complete tracking of the expected current with a small tracking error when the disturbances change randomly and violently.Through the works in this paper,IESO has been proved powerful in uncertainties observation,and it could be combined with multiple algorithms to improve the robustness of the control system,which makes it highly practical.