基于最小方差递推算法的电力系统间谐波谱估计仿真分析

Interharmonic Spectral Estimation Simulation Analysis in Power System Based on Minimum Variance Recursive Algorithm

在对电力系统间谐波进行分析时,提出了基于最小方差递推算法的间谐波谱估计新方法。该方法通过解一有约束的极值问题来求取FIR滤波器的系数,从而得到最小方差谱估计结果,并通过使用Gohberg-Semencul公式和Levinson递推算法使运算变得更为便捷。通过对电力系统间谐波进行仿真分析,同时与MUSIC算法、Burg算法、FFT算法进行比较,表明该算法的稳健性较好,在时变噪声干扰的情况下依然有较高的检测精度,可以满足对电力系统间谐波的分析需求。

This paper presented a new method for analyzing lntemmmu^u minimum variance recursive algorithm. By solving a problem of an extreme value with constraint to obtain the coefficients of F1R filter, then could get the minimum variance spectral estimation by the coefficients. The calculation was faster by running Levinson reeursion and Gohberg-Semeneul formula. By comparing with MUSIC algorithm, Burg algorithm and FFT algorithm, experimental results showed that this algorithm had better robustness, even could achieve high precision in much noisy contaminated conditions, which could meet the demand for interharmonie analysis in power system.