采用自适应线性神经网络(adaptive linear neural network,ADALINE)算法进行永磁同步电机电流谐波检测时,存在电流重构精度不高,在变速、加载阶段电流重构系数与整体信号波形动态变化较大等问题.针对此,提出一种改进型永磁同步电机电流谐波快速重构提取方法:采用d轴和q轴谐波电流提取模块,结合电流瞬态判断单元与增益矩阵自动更新机制,保证高水平电流重构精度,同时提升谐波检测的动态响应特性.仿真实验表明,该方法降低了信号重构对瞬时状态变化的敏感度,电流瞬态判断单元与增益矩阵更新机制加快了电流谐波提取的响应速率,且对存储资源要求低,无需开辟存放大量数据的数组空间.
An Improved Method for Fast Reconstruction and Extraction of Current Harmonics in Permanent Magnet Synchronous Motors
A fast reconstruction and extraction method for current harmonics of permanent magnet synchronous motors is developed to address the problem of low current reconstruction accuracy,especially in the variable speed and loading stages where the current reconstruction coefficient and overall signal waveform change dynamically when using the adaptive linear neural network(ADALINE)algorithm to detect the current harmonics of permanent magnet synchronous motors.The d-axis and q-axis harmonic current extraction modules are used,combined with the current transient judgment unit and the gain matrix automatic update mechanism,to ensure high current reconstruction accuracy and improve the dynamic response characteristics of harmonic detection.Simulation experimental results show that the new method reduces the sensitivity of signal reconstruction to instantaneous state changes.The current transient judgment unit and gain matrix update mechanism accelerate the response rate of current harmonic extraction without creating array space to store the huge amount of data.
万方数据
维普
