The Research on Complex Lithology Identification Based on Well Logs: A Case Study of Lower 1st Member of the Shahejie Formation in Raoyang Sag

Lithology identification is the basis for sweet spot evaluation, prediction, and precise exploratory deployment has important guiding significance areas with low exploration degrees. The lithology of shale strata, which are composed fine-grained sediments, complex varies regularly in vertical direction. Identifying a typical nonlinear classification problem, intelligent algorithms can effectively solve this but different have advantages disadvantages. Compared were three Fisher discriminant analysis, BP neural network, regression decision tree (C&RT) on seven lithologies strata lower 1st member Shahejie Formation (Es1L) Raoyang sag. analysis method linear discriminant, recognition effect poor, accuracy 52.4%; network to identify 82.3%, it belongs black box not be visualized; C&RT accurately Es1L, 85.7%, interlayer thin strata.