密集频率数字信号的判定和校正方法

数字信号的频谱分析中，DFT只得到的频谱可以粗略确定实验信号各谐波频率，振幅和相位，单频谱谐波在其频率的某一邻域内的细化幅值频谱和相位频谱具有显著的特征，通过分析比较，单频率谐波细化频谱与矩形窗的频谱极其相似，依此为准绳，可以判定密集频率信号，进而通过待定谐波参数，选择合适的参数区间和步长组合循环计算，并用矩形窗频谱近似单频率谐波细化频谱的办法，则可以还原校正密集频率的谐波参数，校正精度略低于细化频谱对单频率谐波的计算结果，该方法可以较好的进行情况多变的多个密集频率频谱分析，越多的密集频谱，需要更大的计算量。

Identification and correction for a digital signal with close frequencies

In the process of digital signal processing, DFT can be roughly determined each harmonic frequency, amplitude and phase of experimental signal. For a single harmonic, the amplitude zoom spectrum and phase spectrum have specific features near the harmonic frequency. Through analysis and comparison, a single frequency zoom spectrum and rectangle window spectrum is very similar, and so as the criterion, the digital signal included several frequency harmonic can be determined. Then select the appropriate parameter range and step size to combine for cycle calculation. And make the single frequency zoom spectrum a rectangle window spectrum approximatively. Harmonic parameters can restore from the intensive frequency digital signal. Calculation precision was slightly lower than those of single-frequency harmonic zoom spectrum calculation. The method can better solve the situation variety of two intensive frequency spectrum analyses. More intensive frequencys spectrum analyses need a large amount time to compute.