Kaiser窗三谱线插值电力谐波分析

在非同步采样以及非整数周期截断情况下,快速傅里叶变换会产生频谱泄露和栅栏效应,使计算结果存在较大误差,无法得到准确参数。常用窗函数固定的旁瓣性能制约了已有加窗插值算法的误差修正效果。而凯赛(Kaiser)窗的主旁瓣高度之间的比重可依据需要自由选择,其主旁瓣能量的比例也近乎最大。文中提出基于Kaiser窗的三谱线插值FFT的电力谐波分析方法,推导了基波以及各次谐波的幅值、频率和初相位的插值修正公式。仿真结果表明,所提算法设计灵活且易于实现,在基波频率波动以及白噪声干扰下都有较高的谐波参数估计准确度,验证算法能够有效消除泄露和栅栏效应的影响,提高了谐波分析的准确性。

An approach to electrical harmonic analysis based on Kaiser window triple-spectrum-line interpolation

The fast Fourier transform （FFT） can produce spectral leakage and fence effect under the situation of non-synchronized sampling and non-integral period truncation, which can make a great error in the results of calculation. Common window function fixed sidelobe performance restrains the existing windowed interpolation algorithm error correction effect. However, ratio between mainlobe width and sidelobe width of Kaiser window can be chosen freely according to the need, besides, it can reach nearly the maximum energy proportion of the mainlobe to sidelobe. This paper proposed an approach for electrical harmonic analysis based on Kaiser window triple-spectrum-line interpolation FFT. And the interpolation correction formula of the fundamental and harmonic frequencies, amplitudes and initial phase has been deduced. Simulation results show that the algorithm proposed in this paper can be designed and realized easily, with a higher accuracy of the harmonic parameter estimation when white noise exists or fundamental frequency changes, which verifies the algorithm can eliminate the influence of spectral leakage and fence effect effectively, improving the accuracy of harmonic analysis.