基于人工神经网络的岩石含油气性评价方法  

AN ARTIFICIAL NEURAL NETWORK BASED METHOD FOR EVALUATING THE OIL PROSPECTS OF ROCKS

作　　者：刘勇健[1] 沈军[1] 刘义建 王琳[3] 

机构地区：[1]广东工业大学,广东广州510500 [2]湖南省地勘局407队,湖南怀化418000 [3]中国石化荆州新区勘探研究所,湖北荆州434100

出　　处：《石油实验地质》2000年第3期276-279,共4页Petroleum Geology & Experiment

摘　　要：神经网络计算法是模拟人体经络系统活动的机理来研究事物的新方法。本文运用T .Kohonen提出的SOM自组织人工神经网络模型 ,以准噶尔盆地的彩南油田一主力油层为例 ,建立起岩石含油气性评价的人工神经网络模型。实例研究表明 ,人工神经网络法性能良好 。The calculating method of neural networks is a new way to discuss problems by simulating the mechanism of activities of human body network systems. Based on the SOM artificial neural network model advanced by T. Kohonen, an artificial network model for evaluating the oil prospects of rocks is built in this paper by taking the main reservoirs of Cainan oil field in the Junggar Basin as an example. The case study indicates that the method of artificial neural networks has good performance and is an effective way to evaluate the oil prospects of rocks.

关 键 词：自组织 含油气性 人工神经网络 岩石 评价方法 

分 类 号：TE122.115[石油与天然气工程—油气勘探]