利用方向约束蚁群算法识别断层

Fault identification by orientation constraint ant colony algorithm

断层属性体在断层处表现为属性值的局部极大或极小,如相干体在断层处表现为相干值的局部极小,而且与噪声点不同的是,断层的延伸均有一定的方向。基于此特点,本文提出一种用于断层自动追踪和识别的方向约束蚁群算法,并将该法应用于实际断层识别,通过对比不同参数条件下的自动追踪结果,说明了各个参数的影响。处理结果表明,方向约束蚁群算法是一种准确、有效的断层追踪方法,且在抑制噪声和增强断层连续性方面使用方向约束蚁群算法较Petrel软件具有一定的优越性。

The attribution value at the location of fault generally shows itself as local maximum or minimum.For instance,local minimum indicates the fault in the coherence cube.Furthermore,the fault extension has a certain direction,but noise does not.In this paper,we present a fault automatic tracking and identifying approach named orientation constraint ant colony algorithm(OCACA).We apply the proposed approach in actual coherence cube.Automatic tracking results with different parameters show the influence of the parameters.The processing results demonstrate that OCACA is an accurate and effective approach for fault automatic tracking and identification.And for noise reduction and fault continuity enhancement,OCACA obtains better performances than Petrel.