圆筒型永磁直线同步电机用线性霍尔位置检测的误差补偿

摘要

Position detection based on linear hall sensors for permanent magnetic linear synchronous motor exists error.To tackle this problem, a compensation method combing offline calibration and online Kalman filter was proposed.Errors of hall sensor signals could be classified as DC offset, amplitude imbalance and containing harmonics.The former two factors generate 1st and 2nd order errors in position detection result, and are compensated by offline calibrating use the offset error and amplitude error and ratio achieved by two hall signals.Harmonics in hall sensor signals are caused by air-gap magnetic distortion, the magnetic harmonic content is analyzed by using finite element simulation, coming to the conclusion that magnetic distortion would generate 4th order error in position detection result.So, online Kalman filter was utilized to extract fundamental component from hall signals and to eliminate the influence of harmonics.Experimental results show that the position detection error is reduced by 81%, indicating the validity of the proposed compensation method.%分析了圆筒型永磁直线同步电机采用线性霍尔传感器进行位置检测时存在的误差,并提出了一种离线标定和在线Kalman滤波相结合的误差补偿方法.霍尔传感器信号的误差形式包括存在直流偏置、两路信号幅值不相等和信号中含有谐波等.前两种误差形式分别会在位置检测结果中产生基频和2倍基频误差,对此通过离线标定的方式获得两路信号的直流偏置和幅值比,并补偿到原始信号中.霍尔信号中的谐波由气隙磁场畸变引起,利用有限元仿真分析了磁场中的谐波含量,分析表明磁场畸变会导致位置检测产生4倍基频误差,对此采用在线Kalman滤波提取霍尔信号中的基波分量,消除了谐波分量的影响.实验结果表明,补偿后的位置检测误差减小了81%,从而验证了该补偿方法的有效性.