Research On SRM Neural Network Modeling And Indirect Torque Control Strategy

In today's era of energy conservation and emission reduction,Switched Relutance Motor(SRM)has the characteristics of simple structure,low cost,high efficiency and energy saving,meanwhile,its other features like large starting torque and wide speed range make it found in aerospace,transportation and household appliances.However,there are still many factors that restrict its development.Its special structure makes it highly nonlinear,therefore,building accurate mathematical models has become one of the research directions.At the same time,the traditional PI control used in the SRM speed regulation system is not robust enough to meet the superior speed regulation requirements.Moreover,the torque ripple existing in the actual use of SRM not only generates noise,but also affects the overall system operation,which has seriously hindered the development and scope of SRM.Therefore,this paper analyzes and studies step by step based on the above problems.First of all,based on the SRM indirect torque control system,this paper proposes a modeling method of BP neural network based on nonlinear functions for the problem of SRM's strong coupling and strong nonlinearity.The measured SRM electromagnetic characteristics is employed as sample data,and prior knowledge is fully utilized to select a non-linear function of flux linkage and torque that can initially reflect the non-linear characteristics of SRM to preprocess the sample data.Compared with the traditional BP neural network,the nonlinear BP neural network effectively reduces the number of network nodes,improves the accuracy of SRM modeling and generalization ability,and provides a basis for the accurate implementation of the SRM algorithm.Secondly,for the torque ripple problem caused by the fact that the actual torque cannot track the reference torque in time during the SRM commutation,a full-region compensation torque distribution function is designed.In order to reduce the chattering problem of the sliding mode controller,a new approach law is employed to design the sliding mode controller.At the same time,in order to resist the influence of load disturbance on the speed control performance of SRM,the reduced order expansion state observer is designed in this paper to estimate the load value,so as to improve the ability of SRM to resist load disturbance and the robustness of the system,and realize the high performance in speed control of the SRM system.Then,in view of the problem that the traditional PI control of the outer ring speed regulation ring cannot achieve high performance speed regulation in the face of changes in external conditions,a sliding mode control(SMC)speed control based on the reduced order extended state observer(RESO)is designed.In order to reduce the chattering problem of sliding mode controller,this paper adopts a new approach law to realize the design of sliding mode controller.At the same time,in order to resist the impact of load disturbance on SRM speed regulation performance,this paper designs a reduced-order extended state observer to estimate the load value,so as to improve SRM's ability to resist load disturbance and system robustness,and then realize SRM High-performance speed regulation of the system.Finally,based on the above design,a Simulink simulation model of the SRM indirect torque control system is established,and the simulation experiment is compared with the traditional indirect torque control system from the three aspects in speed change,load change and torque ripple.The validity of the algorithm designed in this paper is verified.