基于支持向量机的致密气藏岩性识别

Lithology identification of tight gas reservoirs based on supporting vector machine

岩性识别是进行测井解释、储层评价的前提。苏里格气田南部上古生界是典型的低孔、低渗致密砂岩气藏．非均质性较强，岩性识别难度大，传统的交会图等常规方法在识别岩性时受到一定的限制。应用主成分分析原理对反映岩性特征的参数进行优化，然后对实际测井资料运用支持向量机方法建立致密气藏岩性数据集．对致密砂岩气层未知岩性进行判别分析，获得了较好的应用效果，岩性识别正确率达76．67％，可以在一定范围内推广。

Lithology identification is the premise of 10g interpretation and reservoir evaluation. Upper Paleozoic of southern Sulige gasfield is a typical tight sandstone gas reservoir with low porosity, low permeability, and strong heterogeneity. It＇s difficult to identify the lithology. The traditional crossplot and other conventional methods have certain restrictions in identifying the lithology. In this pa- per, the principle of the principal component analysis is used to optimize the parameters to reflect the lithological characteristics. Then the tight gas reservoir lithology data sets are established by using support vector machine method on the basis of the actual log data. Unknown lithology of tight sandstone gas reservoirs is analyzed and distinguished, obtaining good application results. Lithology recognition accuracy reached 76.67%, and can be promoted within a certain range.