有源电力滤波器的神经元自适应谐波电流算法研究

随着以电力电子装置为代表的非线性负载广泛应用于电网,电力系统中的谐波污染日益严重。加之用户对电能质量的要求越来越高,探索有效的谐波抑制和无功补偿方法逐渐提上日程。有源电力滤波器是一种新型的能抑制闪变无功和谐波的滤波器,以其优良的性能近年来受到广泛重视。谐波电流检测方法是决定有源电力滤波器(APF)性能的关键因素之一。在传统方法的基础上,结合自适应噪声消除技术,本文提出一种神经元自适应谐波电路检测算法,仿真结果证明该方法鲁棒性强,检测精度高,动态响应好。

A Novel Reference Current Algorithm Based on Neuron Adaptive for Shunt Active Power Filter

With the wide use of non-linear load such as power electronics in the power system, the harmonic pollution has become more and more serious. Given that the customers have raised stricter demand upon the power quality, the effective methods for harmonic suppression and reactive power compensation are more and more needed. Active power filter (APF), a novel filter, able to suppress variable harmonics and reactive power, has attracted more attention based on its excellent performance. Reference current algorithm is a key factor of APF. A neuron adaptive approach for harmonic current calculation is proposed in the paper, according to the adaptive noise canceling technology (ANCT), whose robustness, precision and dynamic response is verified by simulation results.