基于加权一致性的蚁群算法在断层检测中的应用

突出地震数据振幅空间不连续性的断层增强属性体是断层自动解释的基础,而如何压制噪声、地层残余响应、角度不整合接触等伪断层的影响是断层自动检测的关键问题之一.在指纹图像边缘检测处理中,一致性是对指示局部方向场信息各向异性强度的一个度量,本文将其作为一种新的断层增强属性引入到地震数据处理中.为了更有效地压制噪声,本文通过对一致性进行加权处理,再结合蚁群算法,提出了新的断层自动检测方法——基于加权一致性的蚁群算法.理论地震模型和实际地震数据的处理结果表明,对于相同的断层检测过程,对比常规的基于方差属性的蚁群算法和基于C3相干属性的蚁群算法,本文所提方法不仅更有效地压制了噪声和地层残余响应产生的伪断层信息,而且更完整地展现了断层的延伸长度,同时具有更高的稳定性和计算效率.

Application of the ant colony algorithm based on weighted coherence to fault detection

It is the basis of automatic interpretation of faults to highlight the amplitude spatial discontinuity of seismic data,while how to suppress the noise, the formation residual response, the angular unconformable contact, and other pseudo faults is one of the key issues in this approach.Coherence is a measure of the strength of anisotropy that indicates local direction field information in the detection process of fingerprint image edge,which is applied to seismic data processing as a new fault enhancement attribute in this paper.In order to suppress noise more effectively,the coherence is weighted firstly. Then we propose a new fault automatic detection method called the colony algorithm based on weighted coherence,which is formed by the weighted coherence and ant colony algorithm.Numerical tests on synthetic and real field seismic data show that the proposed method in this paper is superior to the conventional ant colony algorithm based on variance attribute and the other based on C3 coherence attribute for the same fault detection process. It can not only more effectively suppress the influence of pseudo fault information generated by the noise interference and formation residual response,but also indicate the extension length of the fault more completely. Meanwhile, this new algorithm has higher stability and computational efficiency than the existing ones aforementioned.