基于AR谱估计和插值FFT的间谐波检测方法

A method for interharmonic measurement based on AR model spectral estimation and interpolation FFT algorithm

现有插值FFT算法是由已知仿真信号频率成分附近的谱线来修正FFT的结果,而实际信号的间谐波和谐波分布往往无法事先确定,这将给插值修正带来不便。提出AR模型谱估计与双峰谱线修正算法相结合的间谐波检测方法。根据信号的AR谱分布进行插值修正,同时提出由谱估计确定B lackm an-Harris窗插值修正所需最小数据长度的方法,并采用多项式逼近的方法导出B lackm an-Harris窗插值算法的简单修正公式,在减少FFT计算量的同时保证了结果的高精度。仿真结果表明了该方法的有效性。

Interpolation FFT algorithms in existence correct results of FFT by spectral lines near the known harmonic frequency of simulated signal. However, the distribution of interharmonics and harmonics in actual signal usually cant be ascertained beforehand, which brings troubles to the application of interpolation and correction. A method for interharmonic measurement is presented based on AR model spectral estimation and correction algorithm of double spectral lines, Interpolation and correction is employed according to the distribution of AR spectra. The method that the minimum length of data needed by Blackman-Harris window interpolation algorithm can be obtained by AR spectra is presented, and the polynomial approximation method is employed to obtain simple correction formula of Blackman-Harris window interpolation algorithm. It can reduce the calculation of FFT and guarantee the high accuracy of results. The simulation results verify the effectiveness of the proposed method.