Research On Adaptive Parameter Identification Method Of Interior Permanent Magnet Synchronous Moter

Interior permanent magnet synchronous motor(Interior Permanent Magnet Synchronous Motor,IPMSM)has the advantages of high power density,small electromagnetic torque ripple,wide speed range,etc.,used in electric vehicles,highprecision servo control and other occasions.Due to the saliency characteristic of IPMSM,the Maximum Torque Per Ampere(MTPA)algorithm that makes full use of its reluctance torque has been widely used.This method can make the system output the maximum electromagnetic torque by adjusting the direction of the stator current vector.However,motor parameters such as inductance and flux linkage will be affected by ambient temperature,magnetic saturation effects,etc.,and will change.This will cause the MTPA’s operating point to shift,making it difficult for the algorithm to track in real time to achieve the best control effect.Therefore,how to implement MTPA current tracking accurately and quickly is an urgent problem to be solved.In response to this problem,this thesis focuses on the parameter identification method of the IPMSM.Focus on the study of using the model reference adaptive method and the high-frequency signal injection method to identify inductance and flux linkage of the IPMSM,to improving the torque-current characteristics of the interior permanent magnet synchronous motor MTPA control system.Model reference adaptive method has the characteristics of fast identification speed,high accuracy,and simple algorithm in the field of parameter identification.Therefore,this paper first constructs the reference model and the adjustable model of IPMSM,and then converts the error equations of the two into an equivalent nonlinear time-varying system.Finally,based on Popov’s superstability theory,interior permanent magnet synchronous motor identification system for flux linkage and inductance is established.The interior permanent magnet synchronous motor MTPA control needs to identify three parameters and the corresponding motor equation is second order,which has the problem of under-rank.To solve this problem,this thesis introduces the high-frequency signal injection method to identify the d-axis inductance.Inject low-amplitude rotating high-frequency signals into the current loop in the control system,the motor responds under the action of high-frequency signal excitation,the relevant parameters of the motor are included in the response.The realtime d-axis inductance parameters can be obtained from the response amplitude of the motor under high-frequency signals through the band-pass filter and the designed synchronous shaft high-pass filter.Therefore,this thesis combines model reference adaptive parameter identification with high-frequency signal injection method parameter identification.Solve the problem of under-rank in the traditional identification process,and verify the effectiveness of the algorithm under the Matlab/Simulink simulation environment.Finally,plan a software and hardware platform based on TMS320F28335 as the main control chip.Explains some of the key circuits in the platform,and depict the program flow of the control software,Finally,the soundness of the online identification algorithm of the system is tested through experiments,and the experimental results are analyzed.