利用人工神经网络优选页岩气有利开发区域  

Optimize the Exploitation Advantageous Area of Shale Gas Reservoirs by Artificial Neural Network

作　　者：王彬[1] 汤勇[1] 勐睿 郭慧玲[1] 王军[2] 刘佳男[3] 

机构地区：[1]西南石油大学油气藏地质及开发工程国家重点实验室,成都610500 [2]西南石油大学石油与天然气工程学院,成都610500 [3]西南石油大学材料科学与工程学院,成都610500

出　　处：《重庆科技学院学报（自然科学版）》2015年第6期33-35,53,共4页Journal of Chongqing University of Science and Technology：Natural Sciences Edition

基　　金：国家"973"计划"超临界二氧化碳强化页岩气高效开发基础"(2014CB239200)

摘　　要：提出一种基于人工神经网络模型优选页岩气有利开发区域的方法。根据美国页岩气储层7种参数与盈亏平衡点(天然气售价)数据,通过网络输入气藏埋深、单层净厚度、热成熟度、有机碳含量、脆性矿物含量和孔隙度等参数,再由网络输出盈亏平衡点,利用人工神经网络建立输入数据与输出数据关系。预测样本数据误差在5%以内,满足工程设计要求,可用于优选页岩气开发区域。In this paper,a new method based on artificial neural network for optimization of the exploitation advantageous area of shale gas reservoirs is presented. According to the USA shale gas data of 7 kinds of parameters and break- even point,the relationship between the input data and output data has been established by artificial neural network. The input data include gas depths,single net thickness,thermal maturity,the content of organic carbon,brittle mineral content and porosity,and the output data is break- even point. The sample data error is less than5%,which meets the engineering requirement,so artificial neural network can be used to optimize the advantageous area of shale gas reservoirs.

关 键 词：页岩气 神经网络 盈亏平衡点 

分 类 号：P618[天文地球—矿床学]