论文部分内容阅读
时窗函数的性质对修正数字滤波器的特性起着决定性作用,所以研究理想的时窗函数显得特别重要。目前,地震资料数字处理中普遍使用的巴特利特时窗函数存在两个明显的问题:①剩余截尾误差按周期性规律衰减;②数字带通滤波器陡度呈非对称性。文中指出,人们在实际应用中提供的带通滤波参数,往往是不对称参数,并与理论上的要求恰好相反,这就加剧了滤波器的频谱波动。文中进一步分析了巴特利特时窗函数的频谱陡度和因子长度等参数与修正后的滤波器的频谱相对波动之间的关系。在分析帕曾(Parzen)时窗函数基础上,作者提出一种新的时窗函数,命名为 WWH 时窗函数。用 WWH 时窗函数修正的数字滤波器,不仅频谱的过渡带较窄,而且在通放带和压制带内的波动也很小,大大改善了数字滤波器的性能。
The nature of the time-window function plays a decisive role in modifying the characteristics of the digital filter, so it is of particular importance to study the ideal time-window function. At present, there are two obvious problems in the Bartlett time-window function commonly used in digital processing of seismic data: ① the residual censoring error decays according to the periodicity; ② the steepness of the digital band-pass filter is asymmetric. It is pointed out in the paper that bandpass filtering parameters provided by people in practical applications are often asymmetric parameters, which is the opposite of the theoretical requirements, which aggravates the spectral fluctuation of the filter. In this paper, we further analyze the relationship between the spectral steepness and the factor length of Bartlett time-window function and the relative fluctuations of the spectrum of the modified filter. Based on the Parzen window function, the authors propose a new window function named WWH window function. The digital filters modified with the WWH time window function not only have a narrower transition band but also very little fluctuation in the passband and the compression band, greatly improving the performance of the digital filter.