Font Size: a A A

Research On Current Distribution And Control Technology Of Hybrid Excitation Synchronous Motor Beased On Neural Network

Posted on:2020-07-26Degree:MasterType:Thesis
Country:ChinaCandidate:Z L ZhangFull Text:PDF
GTID:2392330596479283Subject:Detection Technology and Automation
Abstract/Summary:PDF Full Text Request
Hybrid excited sychronous machine(HESM)is a new type of motor with high power density,high torque density and wide speed range.It has broad application prospects in the field of electric vehicles.This topic combines the requirements of electric vehicles and PMSM control theory and control methods to study the current distribution,current decoupling and system control of HESM:Firstly,aiming at the problem that the reduction of the load capacity of the weak magnetic speed regulation in the high-speed zone,a basic current improved control method considering the d-axis back electromotive force is proposed.The method reduces the excitation current/d-axis current while maintaining the back electromotive force(Ebase),and increases the output torque.The simulation experiments show that the output torque of the high-speed area is raised from the original 2N·m to 3N·m.Secondly,aiming at the nonlinear problem of the proposed basic current improved method,a current distribution approximation algorithm based on BP neural network is proposed.The algorithm uses the function approximation method to simplify the nonlinear equation and simplify the structure of the HESM speed control system.The simulation experiment results have achieved the expected results.Thirdly,aiming at the motor parameter perturbation and the reference value of armature current and excitation current oscillation problem,the BP neural network current distribution controller is used to approximate the current reference value under the condition of parameter stability,which effectively reduces the current reference value oscillation and reduce controller parameter sensitivity.The simulation results show that the oscillation amplitude of the excitation current reference value is reduced to 33%in the low speed zone control;the d-axis current reference oscillation amplitude is reduced to 40%in the high-speed zone control,and the q-axis current reference oscillation amplitude is reduced to 20%,the torque oscillation is reduced to 25%.Finally,aiming at the nonlinear coupling problem in HESM current loop controller,the precise linearization control of HESM is proposed.Using the differential geometry theory of nonlinear systems,the direct-axis current,excitation current and mechanical angular velocity are decoupled into three independent,linear subsystems,which realizes the dynamic decoupling of the current loop in the global range of the speed control system to improve the dynamic performance of the system.Through theoretical analysis and simulation experiments,the proposed HESM control method can solve the problem of that the limitation of load capacity in the weak magnetic control,complication calculation of basic current improved control equation,the current reference value oscillates during motor parameter perturbation and load disturbance and the current loop nonlinear coupling problem,which is suitable for HESM speed control systems.
Keywords/Search Tags:Hybrid excitation synchronous motor, Vector control, Current control, Neural Networks, Exact linearization
PDF Full Text Request
Related items