基于近邻传播聚类与曲线拟合的断层识别

Fault recognition based on affinity propagation clustering and curve fitting

为快速准确地将断层从地震数据中识别出来,提出一种基于近邻传播聚类与曲线拟合的断层识别方法。通过层位的横向端点确定层位不连续点;采用近邻传播聚类算法对层位不连续点进行聚类,属于同一类的不连续点用于确定一条断层;使用曲线拟合方法对每一类层位不连续点进行拟合,拟合的曲线即为断层。该方法可以准确地识别断层且计算量较小,通过与其它方法在实际地震数据上的实验对比,验证了其有效性。

To recognize faults from seismic data quickly and accurately,a fault recognition method based on affinity propagation clustering and curve fitting was proposed.Discontinuous points of horizons were located through the horizontal endpoints of the horizon.The discontinuous points were clustered by the affinity propagation clustering algorithm,and the points of the same cluster were used to determine a fault.The discontinuous points of the same cluster were fitted using the curve fitting method,and the curves were the desired faults.The proposed method can recognize the faults accurately with less computation.By the comparison experiments on the real seismic data with the traditional method,the availability of the presented method is demonstrated.