应用贝叶斯最佳二次型判别函数识别含油气火成岩岩性

 各种火成岩在某些地质条件下有可能成为好的油气储层。近年来,已经陆续在渤海湾以及华北等地发现了经次生改造的侵入岩(煌斑岩、花岗岩)、喷出岩(玄武岩、安山岩)、火山碎屑岩(安山质角砾岩、凝灰岩)等较好的油气储层。因此,根据测井资料应用一种数学方法自动识别含油气火成岩岩性,对寻找有利的油气产层十分重要(欧阳健、王杰英,1988)。采用的识别方法是贝叶斯最佳二次型判别函数(中国科学院计算中心概率统计组编著,1979;肖义越,1982)。

USE OF BAYES BEST QUADRATIC DISCRIMINANT FUNCTION TO RECOGNIZE LITHOLOGY OF OIL-GAS-BEARING VOLCANIC ROCKS

The application of Bayes best quadratic discriminant function to recognize lithology of oil-gas-bearing volcanic rocks on Jin Zhou 20-2 structure in Yiaodong Bay of Bohai Sea.The technique is used to recognize the sample of unknown lithologies to be fitted with a certain li- thology existing in the database on the known various types of volcanic rocks in response to well logs for the purpose of discreminating oil-gas-bearing volcanic rocks.The discrimination accuracy reaches 100%.Thus this method can be used as one of the effective techniques for conventional lithological interpretation.