基于半监督核均值漂移聚类的地震相识别研究

Research on seismic facies identification based on semi-supervised kernel mean shift clustering

地震相识别是根据地震数据内部结构,将之划分为不同的地震相结构单元.本文研究基于半监督核均值漂移聚类的地震相自动识别算法,有效结合了半监督学习和核均值漂移聚类的优势,不用人为给定聚类个数,并且在聚类过程中方便引入少量地震相先验信息,从而有效提升地震相识别的准确性.理论数据聚类展示了该算法对地震相中的多个结构单元识别准确度较高.北海F3实际数据聚类结果表明,本文算法可以得到合理的地震相个数,与其它六种聚类算法的结果相比,本算法划分的地震相结构层次分明且能够区分细小微层.

Seismic facies identification is divided into different seismic facies structural units according to the internal structure of seismic data. This paper proposes a semi-supervised kernel mean shift（SKMS） clustering algorithm which is applied to seismic facies identification, this method can effectively combine the advantages of semi-supervised learning and kernel mean shift. It is not necessary to set the number of clusters, and the accuracy of seismic phase identification is improved by introducing a priori information to guide the clustering process. The clustering results of theoretical data model show that the proposed algorithm is more accurate for the clustering results of several seismic structural units. And the clustering results of real data in Netherlands Offshore F3 Block data also show that the algorithm can get a reasonable number of seismic phases. Comparison SKMS clustering results with other six clustering algorithms, SKMS can divide the seismic facies structure hierarchy distinctly, especially the tiny micro layer.