地震资料处理中随机干扰消除方法分析

Random noise elimination in seismic data processing

由于方法的局限性 ,随机干扰消除技术在生产实践中经常达不到最佳去噪效果。针对这一问题 ,分析了这些方法的基本原理 ,指出了它们的共同前提为有效波具有一定相似性 ,只有满足一定信噪比的地震记录才能使用这些技术。然后将这些方法在形式上统一到矢量分解这一数学原理上。方法的不同 ,主要在于单位相关矢量的求取方式的差异。单位相关矢量的计算方式 ,决定方法在使用上的适应性和局限性。最后给出了理论模型分析与应用实例 。

The random noise elimination methods currently used in practice often can not take advantage of the full potential due to the limitations of each method. Through an investigation of the basic principles behind these methods, the author pointed out that the common premise of these methods is the resemblance of effective signals. Therefore they only work for seismic data with certain extent of signal to noise ratio. These methods then were unified in form with vector decomposition. The difference between each method is how to extract unit correlation vector that restricts the effect of each method in application. Finally some theoretical and real examples were presented, and suggestions on adoption of the random noise elimination methods were proposed.