基于密度的K-means算法在识别含气、含水岩心中的应用
|
摘要
深层火山岩气藏研究一直是地震勘探的难点,火山岩储层固有的岩性和储层空间的复杂性使其油气预测非常困难。针对其特殊性,采用基于密度的K-means算法对深层火山岩含气、含水岩心进行聚类识别。通过分析该算法的聚类识别结果,该算法具有较高的识别准确度与稳定性,因而对地震反演和流体识别具有一定的参考价值。
The research of deep volcanic gas reservoir had been the difficulty of the seismic exploration,it had many difficulties in oil and gas forecast for its inherent lithology and the complexity of the reservoir space of the volcanic reservoir.For its specificity,K-means algorithm based on density was adopted to cluster and recognize the gas-bearing core samples of deep volcanic rocks from water-bearing ones.By analyzed the reoganization results of the algorithm,it has higher accuracy and stability of the clustering recognization,Thus it has some reference value in the seismic inversion and fluid identification.
The research of deep volcanic gas reservoir had been the difficulty of the seismic exploration,it had many difficulties in oil and gas forecast for its inherent lithology and the complexity of the reservoir space of the volcanic reservoir.For its specificity,K-means algorithm based on density was adopted to cluster and recognize the gas-bearing core samples of deep volcanic rocks from water-bearing ones.By analyzed the reoganization results of the algorithm,it has higher accuracy and stability of the clustering recognization,Thus it has some reference value in the seismic inversion and fluid identification.
引文
1唐湘荣.地震测井联合反演方法研究及应用—以塔河4号油区带为例.成都:成都理工大学学位论文,2005
2赖玉霞,刘建平.K-means算法的初始聚类中心的优化.计算机工程与应用,2008;44(10):147—149
3朱明.数据挖掘.合肥:中国科学技术大学出版社,2002:138—139
4闫仁武,雷艳云,任平,等.粒子群的K均值算法在电信客户细分中的应用.科学技术与工程,2009;9(20):6042—6045
2赖玉霞,刘建平.K-means算法的初始聚类中心的优化.计算机工程与应用,2008;44(10):147—149
3朱明.数据挖掘.合肥:中国科学技术大学出版社,2002:138—139
4闫仁武,雷艳云,任平,等.粒子群的K均值算法在电信客户细分中的应用.科学技术与工程,2009;9(20):6042—6045
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心