| Due to the advantage of small size,high power density and high torque density,the permanent magnet synchronous motor(PMSM)has been widely used in industria l applications such as electric vehicles and petroleum.The position sensorless control reduces the size and cost of the system and has robustness,which has become a research hot-spot.However influenced by the inverter nonlinearity and the magnetic spatial harmonics of the motor caused by motor structure nonlinearity,the(6k±1)th harmonics appear in the back electromotive force(EMF),resulting in(6k)th harmonic pulsation in the estimated rotor position obtained by the method based on the back EMF.The presence of harmonics in the rotor position will cause torque ripple and add additional harmonic losses,reducing the operation performance of the PMSM sensorless control.This paper mainly studies the adaptive method to suppress the harmonic error of the estimated rotor position.After analyzing the sliding mode observer based on the back-EMF model,the effects of the inverter nonlinearity caused by the dead time and the magnetic spatial harmonics caused by motor nonlinearity on the back EMF are discussed respectively and the(6k?1)th harmonics in the estimated back EMF exists.The(6k)th harmonics pulsation in the rotor position observation error is derived.The influence of the cutoff frequency of the low-pass filter used in the sliding mode observer and the PI parameters of the quadrature phase-locked loop on the harmonic error in the estimated rotor position are analyzed.It is difficult to determine the appropriate parameters to achieve the purpose of suppressing the estimated rotor position harmonic p ulsation while ensuring the position estimation accuracy and dynamic performance of the system.The correctness of the analysis has been verified by experimental results.A rotor position estimation method based on bilinear recursive least squares(BRLS)adaptive filter is studied.Based on the sliding mode observer,the researched method updates the coefficients of the filter online by using the BRLS adaptive algorithm,so that the filter can adaptively detect and filter out the specified 5th and 7th harmonic components in the back EMF according to the input signal,and then reduce the 6th harmonic pulsation in the rotor position.The BRLS adaptive filter introduces two feedback signals at the same time,which has fast convergence speed and high convergence accuracy.Furthermore,the studied position estimation method can suppress the harmonic error,and the structural design of the proposed method does not depend on the parameters of the PMSM and the inverter.An accurate rotor position estimation strategy based on adaptive fundamental frequency extraction filter(AFFEF)is studied.Based on a sliding mode observer,by using two AFFEFs,the fundamental component of the estimated back-EMF can pass the AFFEF with zero attenuation and zero phase delay,while th e amplitude of the higher harmonic component is seriously attenuated,thereby the purpose of obtaining the fundamental component in the estimated back EMF in real time and filtering out all the higher harmonic components is achieved.Finally,the rotor pos ition is tracked by the quadrature phase-locked loop(PLL).The method is simple in structure,easy to implement,easy to select parameters,and has outstanding steady state performance.The proposed method can effectively suppress the harmonic error in the estimated rotor position and hence improve the estimation accuracy of rotor position.Finally,the reliability of the research accurate rotor position estimation strategy is verified by Matlab/Simulink modeling and the interior PMSM(IPMSM)experimental platform.The experimental results show that the proposed algorithm can effectively compensate the positional harmonic error caused by motor nonlinearity and inverter nonlinearity,and the position estimation accuracy is obviously improved.And the operation performance of the IPMSM sensorless control in high-speed has been improved. |
