功率谱计算（Power spectrum calculation）.docVIP

功率谱计算（Power spectrum calculation） The power spectrum estimation is a very important topic in modern signal processing. Here, with matlab, Ill give you a rough introduction. The power spectrum estimation can be divided into the classical spectral estimation method and the modern spectral estimation method. The simplest of classical spectrum estimation is the periodic graph method, which is divided into direct method and indirect method. The direct method takes the Fourier transform of the N point data, then takes the product of the spectrum and its conjugate, and then gets the estimate of the power spectrum. The indirect method is to calculate the autocorrelation function of the sample data of N point, then take the Fourier transform of the relevant function, which is the estimation of the power spectrum. In matlab, the periodogram method can be implemented using the function periodogram. However, the estimated power spectrum is not fine and the resolution is low. So need to modify periodogram method, signal sequence x (n) can be divided into n not overlapping segments, the periodogram method is used to estimate spectrum respectively, then the results of the n piece of data to estimate the average as the result of the whole piece of data power spectrum estimate. The signal sequence x (n) is also used to calculate the power spectrum and calculate the power spectrum of the data. These two types are called piecewise average periodic graphs, which are generally better than the former. Add window average periodogram method is the improvement of piecewise average cycle diagram method, namely after the data block, for each piece of data preprocessing, plus a nonrectangular Windows and then in the section average period graph method to estimate the power spectrum. Compared to the average periodic chart method, the average periodic chart method can reduce the frequency leakage and increase the width of the frequency peak. The Welch method is to estimate the power spectrum of