Research On The Parameter Identification And Control Method Of Permanent Magnet Synchronous Motortaking Cross Coupling Effect Into Account

Permanent magnet synchronous motor(PMSM)is extensively applied from servo control,aerospace,electic vehicles,to household appliances,due to its simple structure,reliable operation,high efficiency.Whichever maximum torque per ampere(MTPA)control,flux-weakening control or sensorless control is adopted in a PMSM vector control system,accurate motor parameters are needed.In addition,the PI parameters design of the current regulator in control system is also associated with motor parameters.However,affected by temperature,magnetic saturation and other factors,motor parameters often change,and the actual control effect of the system is difficult to achieve the theoretical control effect,therefore,to conduct researches of PMSM parameter identification methods is of important significance.Parameter identification methods and related control methods are researched deeply in this dissertation,and the main work is summarized as follows:Under the influence of magnetic circuit saturation effects and cross coupling effect,the inductance of PMSM will change.The traditional methemetical model of PMSM is established on the condition of ignoring the core saturation,which has difficult to reflect the inductance change.Therefore,a new PMSM methemetical model,taking magnetic circuit saturation effects and cross coupling effect into account,is firstly developed,which lay the foundation of the subsequent researches about motor parameter identification method.Considering the demand of practical application,only a voltage sensor is placed in the DC side of the inverter,and the output voltage of the inverter in AC side is obtained on the basis of the inverter model and drive signals.Therefore,the inverter nonlinear effect must be considered,or the accuracy of the motor parameter identification results will be directly affected.So a compensation strategy of the inverter nonlinear effect,based on phase current-voltage error data table,is proposed,which could realize the compensation of the inverter nonlinear effect using the look-up table during the course of motor operation.In order to obtain the error data table,a special DC experiment is designed.In addition,a novel stator resistance identification method based on unitary linear regression analysis is presented.Using the method,the stator resistance could be calculated according to the reference voltage and actual current recorded in DC experiment.Considering the magnetic circuit saturation effects and cross coupling effect,the sensorless control method of PMSM based on the rotating high-frequency voltage injection has been thoroughly studied.The source of the rotor position estimation error is analyzed theoretically,and a conclusion that the rotor position estimation error is related to d-axis and q-axis dynamic inductance and dynamic cross saturation inductance is made.Then a new dynamic inductance identification method is proposed.Step-by-step identification program is adopted in the dynamic inductance identification method,which could overcome the multi-solution problem due to the deficient rank model.Furthermore,a compensation strategy of the rotor position estimation error is presented.On the basis of the dynamic inductance propotional coefficient has been obtained offline,the dynamic inductance identification,the rotor position estimation error calculation and compensation are achieved online.MTPA control is concerned with d-axis and q-axis static inductances and permanent flux linkage.If the fixed parameter MTPA control strategy is still adopted when motor parameters have changed,the actual control effect will be directly affected.In order to solve the problem,two kinds of static inductance identification methods and a variable-parameter MTPA control strategy are proposed.The influence of the magnetic circuit saturation effect and cross coupling effect is considered by two static inductance identification methods,and the corresponding d-axis and q-axis static inductance at different working point are obtained.The variable-parameter MTPA control strategy could take full advantage of the merit of the per unit system.During the course of motor operation,keeping the torque-optimal current table unchanged,the base value are just updated according to actual motor parameters and the optimal d-axis and q-axis current are obtained using the look-up table,and then more effective MTPA control could be realized.Based on simulation and experiment platform,parameter identification,compensation of the inverter nonlinear effect and the rotor position estimation error,variable-parameter MTPA control have achieved.The feasibility and effectiveness of parameter identification methods,compemsition and control strategies are indicated by simulation and experiment results.