基于渐消记忆递推最小二乘法的空间矢量控制APF参数在线辨识

Parameter identification online of space vector control APF based on fading memory recursive least squares

为抑制三相并联有源电力滤波器在空间矢量控制中因参数变化对补偿效果的影响,本研究采用基于递推最小二乘法对参数实时在线辨识,并将渐消记忆的指数窗口引入递推最小二乘法中,克服了递推最小二乘法出现的"数据饱和"现象。仿真结果表明该方法简单可靠,鲁棒性强,易于离散化实现,能很好在线辨识APF输出等效阻感,改善了空间矢量控制性能。

To suppress the effect on the compensation when the parameter changes in three-- phase shunt an active power filter in space vector control, a method is used based on real--time online recursive least squares method parameter identification. And introducing the fading memo- ry into the exponential window which overcome the appearance of ＂Data Saturation＂ phenomenon in the recursive least squares method. Simulation results show that the method is simple and reli- able, robust, easy to implement discrete, and well identifies the APF output equivalent resistance online, and improves the performance of space vector control.