Researches On Online Parameter Identification Strategy Of IPMSM

The interior permanent magnet synchronous motor(IPMSM)is widely used in electric vehicles and has many advantages,such as high torque density.However,due to the influence of magnetic saturation and cross-coupling,the motor parameters present strong nonlinear characteristics in practical applications.The maximum torque per current(MTPA)control strategy is usually used in the IPMSM.The efficiency optimization results of the strategy will be influenced when we use fixed parameters to calculate the MTPA operating point.Therefore,it is very necessary to develop an effective online parameter identification strategy and update the motor parameters in real-time.In this paper,the existing problems of the IPMSM parameter identification technique are researched.Firstly,the IPMSM electrical model is rank-deficient which means only part of the parameters can be identified synchronously.After analyzing the influence of the parameter variations on the control strategy and the disadvantages of the full-parameter identification scheme,a full-rank parameter identification model is established by setting part of the parameters as fixed values.Secondly,choosing a suitable system identification algorithm for the identification of motor parameters.Considering the existing harmonics of the control system and the calculation burden of the microcontroller,a normalized least mean square(NLMS)based parameter identification model is established.The NLMS algorithm has good filtering performance,moderate convergence rate,good stability,and low calculation burden.Thirdly,the inverter nonlinear factors will cause voltage deviations between the reference voltages and the actual voltages.The voltage deviations influence the accuracy of the parameter identification results,especially under low load conditions.Due to the small weight of estimation,a large error will occur between the identification results and actual parameter values.In this paper,the measured torque is used to calibrate the threshold voltage and conduct resistance.After that,the deviation can be calculated.The accuracy of parameter identification is greatly improved in this way.Fourthly,in the position sensorless control applications,due to the estimation error of the rotator speed,the accuracy of the parameter identification is significantly influenced.In this paper,the reason that causes rotator speed estimation error is analyzed.The accuracy of speed estimation and parameter identification is improved by considering the inverter nonlinear factors.Fifthly,in the control applications of using a position sensor,due to the inaccurate installation position of the sensor,an error exists between the feedback electrical angle and the actual electrical angle.The angle error will influence the FOC control and parameter identification results.Therefore,a method of calibrating the angle position error is developed by simulation modeling analysis.Finally,the experimental platform for the parameter identification of IPMSM is set up and two different experiment groups are designed to verify the proposed parameter identification strategy.All the two experimental results show that the proposed IPMSM parameter identification strategy can greatly improve the accuracy of parameter identification results by considering the influence of non-ideal factors such as voltage deviation caused by inverter nonlinear factors.