有源电力滤波器参考电流的预测方法及其实现

为减小系统及数字化控制器延时对有源电力滤波器(active power filter，APF)补偿性能的影响，基于一种改进的自适应预测方法，该文提出3种APF 参考电流的预测策略，并对它们的稳态预测误差和动态跟踪性能进行实例比较和仿真分析，得出：首先对输入信号进行预处理，再利用其来预测电网基波电流的预测策略是最优的。该文还提出利用自适应预测算法和内插值算法相结合来实现对APF参考电流预测的思想，这样既达到了较高的预测精度，又保证了系数的自适应调整运算有充裕的时间完成，使滤波器系数的自适应调整运算不影响功能程序的执行。所提出的预测控制方法在某铜箔厂注入式混合型有源电力滤波器(injection hybrid active power filter，IHAPF)中进行现场实验，之后成功投入到工业应用。实验结果和工业应用效果均表明文章所提方法的有效性和可行性。

Prediction of the Reference Current of Active Power Filter and Its Realization

Based on a mended adaptive prediction method, three kinds of prediction strategies of the reference current of active power filter (APF) were proposed for diminishing the time-delay effect of digital controller. Their steady state estimation errors and dynamic tracking performance were compared through example and analyzed by simulation. Then the conclusion that predicting fundamental current according to the input current pretreated is the best prediction strategy was reached. Finally, this paper combined adaptive prediction method and internal algorithm of interpolation to realize the prediction of the reference current of APF. In this way, high prediction precision can be reached and the adaptive adjustment of coefficients would have enough time to be completed also and the adjustment program would not affect other executive program. The prediction method proposed in this paper was tested in the injection hybrid active power filter (IHAPF) of one copper plant and has been applied successfully. This confirmed its validity and feasibility.