有源电力滤波器的箕舌线变步长LMS谐波检测方法

将单相有源电力滤波器（APF）作为研究对象，对其负载非线性电流中的谐波电流检测方法进行了进一步的研究．基于自适应噪声对消技术，将基于箕舌线的变步长最小均方（LMS）算法应用于APF谐波检测，同时将其与定步长LMS算法、基于s函数的变步长LMS算法，以及目前最常用的谐波检测算法之一的离散傅里叶滑窗谐波检测方法进行仿真比较．通过仿真分析，该算法具有较小的运算量、稳态误差小、强跟踪能力等特点，并有较好的收敛速度，解决了收敛速度和稳态误差之间的矛盾．

Versiera variable step LMS algorithm for APF harmonic current detection

The methods of the harmonic current detection which are used in the load nonlinear electric current when the single-phase shunt APF is taken as the research object should be made a further research. Versiera variable step least mean square based on the adaptive noise cancellation technology is used to detect APF harmonics. The paper compares the method to fixed step least mean square and variable step least mean square based on the S function and the sliding window harmonic detection algorithm based on Fourier analysis that is one of the most widely used through simulation analysis. The algorithm reduces the amount of calculation furthermore it has small steady state error and a good performance of tracing. This algorithm has a better convergence speed through the simulation analysis, it solves the contradiction between convergence speed and steady state.