摘要
磁链观测法是永磁同步电机在中高速区域一种有效的无传感器控制方案。但是由于逆变器非线性、初始位置误差、电流测量偏置等因素,磁链估计结果将不可避免地含有斜坡信号、直流偏置和各次谐波等多种干扰信号。这些干扰信号的存在将降低无传感器控制的稳态精度,增加转速波动,严重时甚至使系统失稳。针对这一问题,提出了基于扩张式主从自适应陷波滤波器(EMAF)的磁链观测方案。利用不同干扰信号的特性,EMAF可同时跟踪磁链观测中的多种干扰信号;通过对积分对象中的直流偏差进行反馈补偿,EMAF可消除积分环节产生的斜坡信号。进一步地,给出了EMAF的参数整定图解法,将多参数整定问题转换为单参数整定。另外,为了提高频率跟踪环节的动态性能,提出一种考虑电机运动方程的动态频率跟踪算法。相比锁相环或锁频环,该算法在频率变化时可减小频率跟踪误差,避免频率突变引起的暂态振荡。不同条件下的实验结果验证了所提方法的有效性。
The model-based sensorless control of permanent magnet synchronous motor(PMSM)is effective at medium and high speed regions.Among several model-based methods,the flux observation method can achieve better performance because the rotor flux of PMSM is constant or calculable even when the speed or current is time-varying.However,due to the inverter nonlinearity,the initial position detection error,and the current measurement bias,the estimated flux will inevitably contain interfering signals such as ramp signal,DC bias,and harmonics.These interfering signals will reduce the steady-state accuracy of rotor position estimation,increase speed fluctuation,and even make the system unstable in severe cases.Therefore,a flux observer based on an extended master-slave adaptive notch filter(EMAF)is proposed.The EMAF can track the signals containing multi-components by using their characteristics.Thus,considering the characteristic of interfering signals in estimated flux,we design the EMAF,which can accurately track these interfering signals.Besides,by introducing a new variable in EMAF,the dc deviation of integral terms is tracked and compensated,and the ramp signal caused by integrating is also eliminated.This method can directly estimate the rotor flux without estimating the back-EMF first.Since many parameters in EMAF need to tune,a parameter tuning graphical method of EMAF is given to simplify the tuning process,transforming the multi-parameter tuning into single-parameter tuning.In addition,under varying speeds,the traditional phase-locked loop(PLL)can not accurately track the frequency,which degrades the dynamic performance.Therefore,a dynamic frequency tracking algorithm(DFT)is proposed.In this method,the motion equation is applied,and the extended state observer(ESO)is designed to estimate the frequency.Since the varying rate of the rotor speed is considered in ESO,compared with the phase-locked loop or frequency-locked loop,this method is more applicable under changing speed.Based on EMAF and DFT,the overall sensorl
作者
葛扬
宋卫章
杨洋
Ge Yang;Song Weizhang;Yang Yang(School of Electrical Engineering Xi’an University of Technology Xi’an,710061,China)
出处
《电工技术学报》
EI
CSCD
北大核心
2023年第14期3824-3835,共12页
Transactions of China Electrotechnical Society
基金
国家自然科学基金(51877176)
陕西省教育厅服务地方专项项目(18JC024)资助。
