莱夫–文森特窗插值FFT谐波分析方法

加窗插值快速傅里叶变换(fast Fourier transform，FFT)算法广泛应用于电力系统谐波分析，可改善因非同步采样和非整数周期截断造成的频谱泄漏，提高谐波参数计算的准确度。该文分析莱夫–文森特(Rife-Vincent)窗的频谱特性，提出基于5项Rife-Vincent(I)窗插值FFT的谐波分析算法，运用多项式拟合求出简单实用的插值修正公式，大大减少了谐波分析时的计算量。仿真结果表明，在非同步采样和非整数周期截断条件下，提出的谐波分析方法适合于弱信号分量的提取和复杂谐波信号的准确分析，对含21次谐波信号分析的频率计算误差仅为1.9× 10-8%，幅值、初相位计算误差分别小于等于0.000 1%和0.029%。

An Approach for Harmonic Analysis Based on Rife-Vincent Window Interpolation FFT

Widely used for harmonic analysis in the electric power system, the windowed interpolation fast Fourier transform (FFT) algorithms could compensate the spectral leakage in frequency domain caused by non-coherent sampling and non-integral period truncation, and subsequently the accuracy of harmonic parameters measurement could be improved. The spectral characteristics of Rife-Vincent window were analyzed and an approach for harmonic analysis based on five term Rife-Vincent(I) window interpolation FFT was proposed. The applicable rectification formulas of the interpolation were obtained by using polynomial curve fit functions, and subsequently calculating burden was drama- tically reduced. The simulation results indicate that the approach presented in this paper is adapted for the extraction of weak signals components and the accurate analysis of complex harmonics, and by using the approach in the non-coherent sampling and non-integral period truncation conditions, the errors of calculating frequency of 21 orders harmonics is 1.9×10-8%, as well as that of calculating amplitudes and phases are no more than 0.000 1%、0.029%.