滑动DFT算法在功率谱估计中的应用

Application of the sliding DFT algorithm in power spectrum estimation

基于滑动DFT算法推导出一种改进的周期图功率谱估计方法,并在软件系统界面中应用。根据传统的功率谱估计方法和滑动DFT算法推导出改进的功率谱估计算法,通过滑动DFT算法计算出DFT值,计算DFT时通过加窗减少了频谱泄漏,然后通过周期图法计算出最终的功率谱值,在Matlab中绘制出功率谱进行验证,用C++语言对该算法进行实现,导入原始数据进行功率谱估计计算并在界面中绘制出频谱。实验表明,使用滑动DFT算法计算频谱,所需样本点较直接计算FFT更少,计算量较直接计算FFT更小,封装成类后可应用于其他模块中。

An advanced periodogram method was proposed based on the sliding DFT algorithm, which was also applied in the interface of a software system. In this paper, an advanced periodogram method was deduced according to the conventional power spectrum estimation method and the sliding DFT algorithm, using the sliding DFT algorithm and by windowing in order to reduce the spectral leakage when calculating DFT, then calculated the power spectrum by using periodogram method, plotted in Matlab to verify, implemented this method with C＋＋, then imported the raw data and plotted the spectrum in the interface. The results indicated that periodogram method based on the sliding DFT algorithm required fewer sampling points and calculating works. After encapsulated, it can be applied to other modules.