Petroleum reservoir and non-reservoir discrimination analysis based on data mining algorithm

Abstract

Data Mining is an effective tool to “the nontrivial extraction of implicit, previously unknown, and potentially useful information from data”. And in petroleum industry there are numerous data needed to be analyzed to assist the petroleum exploitation decision. In this paper, petroleum reservoir and non-reservoir characteristics reorganization analysis are put forward based on data mining method. Relative date sets collected from petroleum exploration are transformed and integrated first. Three kinds of classification methods are introduced, which are Linear Discriminate Analysis (LDA), Decision Tree and Multi-criteria Linear Programming (MCLP). And they are used here for reservoir and non-reservoir discrimination analysis. The experiment result shows that it is feasible to predict the reservoir level and non-reservoir level in oil field based on the existing history data sets with data mining method and algorithm.