Abstract:The characteristic of time-window functions plays a decisive role in revising the characteristic of digital filter; therefore, it is particularly important to study ideal time-window function, Bartlett timewindow function widely used in present digital seismic data processing has two obvious problems;(1) residual truncation error attenuates periodically; (2) the gradient of digital bandpass filter is asymmetric.It is pointed out that the bandpass filtering parameters people use in practice are often asymmetric and even contrary to theoretical requicements, which aggravates the fluctuation of frequency spectrum of a filter, The paper makes a detailed analysis of the relationship between the parameters of Bartlett time-window function (such as frequency speetrim gradient, operator length, etc, )and the relative fluctuation of frequency spectrum of the filter that has been revised.On the bas is of the analysis of Parzen time-window function,the author put forward a new time-window function called WWH time-window function. The digital filter revised with WWH function brings not only narrow transition band of frequency spectrum but also very low fluctuations in both passband and suppression band.As a result, the characteristic of digital filter is greatly improved.
收稿日期: 1988-01-07
引用本文:
王卫华. 数字滤波器的修正方法分析[J]. 石油地球物理勘探, 1988, 23(5): 576-589.
Wang Weihua. An analysis of the method for digital filtcr revigion. OGP, 1988, 23(5): 576-589.