Research On Global High Efficiency Control Techniques Of Variable Flux Memory Machines

Variable flux memory machine(VFMM)can be regarded as a flux-adjustable machine in real sense.The low coercive force magnets in VFMMs can be magnetized or demagnetized simply by a current pulse.Then the current required by the conventional flux weakening(FW)method at a high speed can be effectively reduced.Thus,VFMM usually has an extended speed range with high efficiency,which makes it become a promising candidate for electric vehicles,and etc.This dissertation focuses on the global high efficiency control techniques of VFMM,and the main research works are summarized as follows:(1)A nonlinear model based on measured parameters for VFMM is established by combining the mathematical model of conventional permanent magnet synchronous machine(PMSM)and the MS manipulation characteristics of VFMM.The characteristics of MS manipulation are obtained by the no-load back electromotive force(back-EMF)test,and the characteristics of flux linkage during the MS manipulation process are obtained by the voltage pulse injection method.The established model is then evaluated by steady-state torque comparison and voltage comparison during the MS manipulation process.(2)An MS initialization method for VFMM without position sensor is proposed.The highfrequency square wave pulse injection method and the bipolar voltage pulse injection method are utilized to estimate the initial rotor position and detect the rotor polarity respectively.The voltage pulse amplitude for the polarity detection is analyzed.The estimated rotor position is used to apply a magnetizing current pulse so as to realize the MS initialization,which can provide the maximum starting torque for the machine.The voltage limitation,torque fluctuation,and loss characteristics during the MS manipulation process are analyzed.The voltage and loss characteristics of different sinusoidal current pulses are compared.An asymmetric current pulse design method is proposed to reduce the losses during the MS manipulation process.(3)A maximum torque per ampere(MTPA)control method using operating region boundary based MS manipulation is proposed,which can reduce the MS manipulation frequency by separating the magnetization and demagnetization boundary curves.Besides,the execution time required by the controller to implement the algorithm is further reduced by using a linear approximation method.Two MTPA control methods based on linear approximation and iterative technique respectively are compared in terms of their execution time,the operating region,and the output current.Experiments are carried out to verify the effectiveness of the proposed two MTPA control methods.(4)A mathematical model for VFMM under uncontrolled generator fault(UCGF)is established,and the operating characteristics of the machine under the fault state are analyzed.Based on the maximum voltage characteristic during the fault,an MS selection method for the prevention of UCGF is proposed.The MS manipulation is executed according to real-time speed to make the maximum voltage during the fault below the allowable voltage limitation.Thus,the damage caused by the UCGF can be completely prevented.(5)A test platform for large-scale VFMM prototype is established,and the hardware and software of the machine driver are developed.According to the operating characteristics of the VFMM,an efficiency test method is proposed.Lookup tables from the test are used to realize the optimal efficiency control of the machine within the entire speed range.The characteristics of MS manipulation,output torque,and efficiency distributions are tested by the established test platform and evaluated by the developed driver.Intensive studies are carried out in this dissertation in terms of mathematical modeling,MS manipulation,and MTPA control for VFMM,which will provide some theoretical and technical support for its industrialization.