Study On Permanent Magnet Synchronous Machine Sensorless Vector Control Technology

Permanent magnet synchronous machine(PMSM)has been widely used in many fields,including aerospace,electric vehicle and industrial control,due to its simple structure,high efficiency and large torque density.The sensorless PMSM vector control technique is further studied in this dissertation to improve its reliability and expand application areas.The middle and high speed control methods based on sliding mode observer(SMO)and disturbance observer as well as the low speed control method based on high frequency signal injection for PMSM are studied in this dissertation.The speed and position estimation for surface-mounted PMSM(SPMSM)based on disturbance observer is proposed by combining with load torque rejection control.The full speed sensorless control method for the interior PMSM(IPMSM)is presented by combining the high frequency signal injection method and the disturbance observer based method.The main studies and achievements are introduced as follows.In order to alleviate the chattering problem in conventional SMO,this dissertation proposes a soft switching SMO for sensorless SPMSM,in which the traditional switching function is replaced with a boundary-layer-flexible sinusoidal saturation function.Equivalent convergence interval is defined so as to get insight into the convergence velocity of different saturation functions,based on which the equivalent convergence time of the sinusoidal saturation function is deduced from sliding mode state between the boundary layers,and compared with the existing elementary,terminal sliding-mode,hyperbolic tangent and exponential saturation functions.The effects of the saturation function types,switching gain,equivalent convergence coefficient and boundary layer thickness on estimation accuracy and convergence time are analyzed according to simulation results.This dissertation proposes a back electromotive force(EMF)observer for SPMSM based on disturbance observer,which is capable of estimating the system disturbance accurately without calculating the state vector derivation or inserting noise filter.A sensorless control method for SPMSM drive is introduced,which is robust against load torque variations.The load torque observer is also constructed based on the disturbance observer method to compensate for load torque disturbance.Considering the estimated speed is served as the state vector in the establishment of the load torque observer,the gains of the two observers are carefully selected to improve estimation accuracy and dynamic performance.The simulation results verify the effectiveness of the sensorless method and the adjustment laws for the proposed observer gains.A full speed sensorless control method is presented for IPMSM drive.An extended EMF observer is proposed based on a deliberately designed disturbance observer to estimate the rotor position.The proposed disturbance observer based position extraction method is combined with the pulsating HF injection method to achieve high-performance sensorless control of IPMSM in the full speed region.In order to achieve high-efficiency operation,a parameter-irrelevant maximum torque per ampere(MTPA)control scheme is proposed based on the perturb and observe(P&O)method.A hardware platform for the sensorless PMSM control is constructed to experimentally verify the sensorless methods proposed in this dissertation.Experimental results show the optimal SMO is effective for the speed and rotor position extraction and confirm the ability of chattering elimination.The experimental results also demonstrate that the disturbance observer based sensorless SPMSM control system is self-adjustable with speed and robust against load variation.Furthermore,the results of evaluative experiments verify the effectiveness of the proposed high-performance sensorless control and torque disturbance rejection for SPMSM drive.The experimental results validate the effectiveness of the proposed full speed sensorless method and the P&O based MPTA control algorithm for IPMSM drive.