Research On Sensorless Control Of Interior Permanent Magnet Synchronous Motor Control System

Electric vehicle(EV)industry is developing rapidly because of the presure of energy saving and environment protection.The driving system of an EV usually consists of permanent magnet synchronous motor(PMSM)using vector control algorithm.The precise rotor position needs to be measured by mechanical position sensor during the motor running period.However,additional sensor increases the cost and complexity of the motor dri ving system,and the adaptability of position sensor is limited by some extreme conditions.Therefor e the sensorless control stratagies of PMSM have been widely studied.This paper focus on the sensorless control algorithms of interior permanent magnet synchronous motor(IPMSM)which is more applicable to EV s.Firstly,the ideal mathematical model of IPMSM is introduced.An IPMSM control system simulation platform of which the motor parameters can be changed during simulation time is built with MATLAB/Simulink.The further research of sensorless control methods are all based on this platform.We test the new built signal IPMSM model with the model that the Simulink offers,and the response curves shows the new model is reliable.The PI parameters of speed loop and current loop are adjusted properly,and the speed and torque response of IPMSM is tested.And an evaluation method of rotor position error is proposed.Secondly,we study two sensorless control algorithms for medium and high speed region of IPMSM,one is model reference adaptive system(MRAS)method,another is sliding mode observer(SMO)method.The theories of both algorithms have been introduced.We choose the Popov hyper-stability theory to help designing the nonlinear time-varing feedback system that is equal to the MRAS,then we prove the hyper-stability of the system,and the adaptive law to estimate rotor speed is designed.We design the SMO based on the extended back electromotive force,and study the necessary condition of the gain parameter of the switch control function of SMO.In the whole speed rigion,the speed,torque response and rotor positon estimation error of these two sensorless control methods have been tested with the above-mentioned simuation platform.And the sensitivity of esitimated position to the variation of different IPMSM parameters is studied.For MRAS method,we find the variation of different parameters has different impacts on the estimated position error in one speed region,and in different speed regions,the variation of one parameter also has different impacts.In whole speed region,the flux linkage of rotor influences the most,and the stator resistance influences much more in low speed rigion compared to high speed conditions.For SMO method,the identified position error is most sensitive to the variation of q-axis inductance in the whole speed range.The robustness of estimated rotor position to the variation of parameters of SMO sensorless control is stronger than that of MRAS.Thirdly,we study the high frequency sinusoidal injection(HFSI)sensorless control method for the low speed region of IPMSM control system.The high frequency current response with rotor position information is deduced in both rotating HFSI and pulsating HFSI methods,and the Luenberger observers are designed to estimate the rotor position.From the simulation results,we can find the estimated rotor position is very accurate with these two HFSI methods under low speed and no load torque condition.