有源滤波器电流跟踪控制的一种新方法

在采用有源电力滤波器(APF)消除谐波的过程中,由于滤波器的检测精度有限、指令电流计算延时和输出滤波器相移等因素的影响,传统的PI控制滤波效果不够理想.针对APF电流跟踪控制问题,提出了一种基于PI型学习律的迭代学习控制算法,通过实施遗忘因子学习律增强系统鲁棒性的同时,引入了在最优目标选择下模糊参考电流误差的D型学习律前馈环节,提高了系统的跟踪精度.仿真与现场运行结果都证明了该控制方法具有谐波电流跟踪效果好、计算量小、易于工程实现等特点.

New Method for Current Tracking Control of Active Power Filter

Due to the influences of detection precision, time delay of instruction current calculation and phase displacement of output filter, harmonic elimination using active power filter (APF) with traditional PI controller is not ideal. A Pi-type iterative learning control algorithm for APF current-tracking control is presented, in which, a forgetting factor is used to enhance the systemic robustness, and a feed forward using D-type learning law based on the fuzzy-inferring referenced current error under the optimal object is introduced to improve APF' s dynamic performance. The proposed algorithm has a good tracking effect, a small computation capacity and is easy for engineering implementation. The simulation and running results all validate its value.