周期信号频谱分析应注意的问题

Some Problems About the Spectrum Analysis of Periodic Signal

在实际应用中，常用DFT算法对周期信号作频谱分析，但周期信号的频谱和一般信号的频谱有着不同的物理意义。一般信号的DFT结果表示该信号在频率Ω处的频谱密度，它是一个连续信号，周期信号的DFT仅表示它的基波及各次谐波的幅值，是一个离散信号。一些对一般信号进行频谱分析的方法，使用于周期信号则不合适。另外，如果采样时基选择不好、时基偏移或干扰等原因，都会造成泄漏，给频谱分析带来混淆。主要讨论周期信号频谱分析的特点及应注意的问题。

In practice, DFT algorithm is usually applied on the spectrum analysis of periodic signals. The spectrum analysis of periodic signals has some difference with the spectrum analysis of nonperiodic signals. The DFT result of nonperiodic signals indicates the spectrum density on frequency Ω, and is continuous, The DFT result of periodic signals indicates the amplitude of harmonic and is discrete. If sample timereference not selected properly, or sample the drifted, or other disturbance, all can cause spectrum leakal and get a wrong analysis result. The characteristic about the analysis of periodic signals and some noticed problems are discussed.