The Identification Method of Igneous Rock Lithology Based on Data Mining Technology

Abstract

Igenous rock is featured with complex and multivariant lithology, its logging response is multiplicity, therefore, it is difficult to identify igneous rock lithology with logging data. To Lay the foundation for fine logging evaluation of igneous reservoir in Songnan gas field, the identification method of igneous rock lithology is researched. Common crossplot method identify lithology with only two logging parameters, its precision is not high. To improve the accuracy of identification, the data mining software, named as weak, is used, three data mining methods, including Association Rule, Decision Tree, Support vector machine, are applied in lithology identification. The results show that these methods can improve the accuracy of lithology identification, in particular, Decision Tree model has the highest recognition accuracy, while it is relatively easy to understand, so it can be used as auxiliary tools for recognition of igneous rocks. Decision Tree model is used to process logging data of exploratory well, and its computation results are well consistent with the core thin section data.