Research On The Position Sensorless Hybrid Control Strategy For PMSM

Permanent magnet synchronous machines(PMSMs)have been widely used in electric drive systems.High performance PMSM drives usually adopt the vector control strategy in which the knowledge of the rotor position is a prerequisite.A promising alternative to replace the problematic mechanical position sensors is the sensorless control technique.The elimination of the mechanical position sensors can reduce the overall cost and volume,improve the power density and the system reliability and also make it easy for assembling and cabling.This thesis presents a position sensorless hybrid control strategy for the PMSM,aiming at improving the position and speed estimation accuracy in the whole speed range.The proposed strategies can be applied in the PMSM drive systems of home appliances like washing machines and air-conditioners,or in the drive systems of electric vehicles.In the proposed strategy,an improved high-frequency voltage injection(HFVI)method is proposed for the sensorless operation in the standstill and low-speed region,while an improved method based on the sliding mode observer(SMO)is proposed for the position and speed estimation in the middle-and high-speed region.The innovation of the improved HFVI is twofold.Firstly,the use of low-pass filters(LPFs)is replaced by the characteristic harmonic elimination(CHE)method,thus avoiding the phase delay and estimation error problem brought by the LPFs.Secondly,the analysis of the new method considers the stator resistance,which is usually neglected by the conventional methods,and the compensation method is proposed accordingly to improve the estimation accuracy.In the middle-and high-speed region,new methods are proposed to alleviate the chattering problem in the proposed SMO.To extend the speed range of the drive,the maximum torque per Ampere(MTPA)control strategy is adopted below the base speed,and the flux-weakening(FW)strategy is adopted above the base speed.The inductance variation is considered in both strategies.The key techniques for the smooth switching between the HFVI and SMO methods as well as the smooth switching between MTPA and FW are described in detail.To testify the proposed sensorless control scheme,simulations are carried out in the Matlab/Simulink environment and experiments are performed in an industrial PMSM drive system.The starting from standstill,response to the step-changed reference speed,response to step-changed load and efficiency test are included.The results show that the position and speed estimation accuracy is improved by the proposed methods.Good steady-state and transient performance of the closed-loop sensorless operation is achieved.Stable starting of the machine and smooth switching between the two sensorless techniques are realized.The feasibility and advantages of the proposed position sensorless hybrid control strategy are testified.The proposed strategy can be applied to PMSM drives of home appliances and electric vehicles.