离散傅里叶变换是对电力系统稳态信号进行频谱分析的最基本数学工具,也是国际电工委员会推荐用于谐波和间谐波测量的变换方法。该方法在分析窗口长度与实际信号周期不符时,各频率成分间会发生频谱干扰,从而产生较大的分析误差。对此,提出一种基于余弦组合窗的多层插值频域校正法,用于电力系统谐波分析。该方法利用旋转调整后各离散谱线的相位特点,通过多层求和计算,使各非关注成分在各关注成分对应谱线上的泄漏影响达到最小,因此,其能够在加窗的基础上,进一步抑制信号间的频谱干扰。仿真算例表明,该方法能够在非整周期采样的条件下实现电力系统谐波信号的高精度测量,也从另一个角度改进了传统的加窗插值算法。
Discrete Fourier transformer is the most basic mathematical tool for spectrum analysis of a stable electrical signal, which is also the transformation approach recommended by International Electro-technical Commission for harmonic and interharmonic measurement. However, when the analysis window width does not correspond to the actual signal period, spectral interferences among different tones would take place by using the above method, thus making the analysis accuracy deteriorated. To address this problem, this paper proposed a multilayer interpolation correction approach based on cosine windows for power system harmonic analysis. This method uses the phase features of the spectral bins after rotation adjustment to make the interferences on the concerned tone, caused by the unconcerned tones, minimum through multilayer summation. It can further restrain the spectral interferences among different tones on the basis of windowing. The simulation case study shows that the proposed approach can achieve high-accuracy harmonic measurement even when the sampling does not exactly fit the integral signal period. It improves the traditional windowed interpolation approach from another point of view.