应用量子神经网络预测低渗储层水锁损害  

Prediction of Water-locking Damage of Low Permeability Reservoirs by Using Quantum Neural Network

作　　者：孙玉学[1] 谢建波[1] 才庆[2] 

机构地区：[1]东北石油大学,黑龙江大庆163318 [2]中油大庆油田有限责任公司,黑龙江大庆163453

出　　处：《特种油气藏》2012年第6期53-55,143,共3页Special Oil & Gas Reservoirs

基　　金：中油大庆油田有限责任公司项目"古龙西地区储层损害机理及保护措施研究"(Dq-1204003201-js-318)

摘　　要：对于低渗储层最主要的损害类型水锁损害进行及时、准确的预测,在油气层保护中起着至关重要的作用。在分析水锁损害产生机理和各种影响因素的基础上,基于生物神经元对信息处理方式和量子神经算法原理构造出一种量子神经元,建立预测储层水锁的量子神经网络模型,并编制软件,进行气藏水锁损害预测。该方法克服了灰关联分析法需要进行评价矩阵分析、操作复杂的缺陷,经大庆油田龙西地区实践证明,该方法运算速度快,系统所需参数少,准确率高(总体符合率达到90%),可为低渗储层保护技术提供可靠的支持。Timely and accurate prediction of water-locking damage of low permeability reservoirs plays an important role in reservoir protection.The paper,based on the analyses of the mechanism of water-lock and the factors affecting water-lock,raises a quantum neural network unit based on information processing mechanisms of biological neuron and the calculation method of quantum neuron;the paper also establishes quantum neural network model to predict water-lock;and a software to predict water-lock is programmed.The method overcomes the defects of the grey relational analysis method that needs evaluating matrix analyses and complex operation.The application of the method in Xilong area of Daqing Oilfield shows that the method has got the advantages of fast calculation,small parameters required and high accuracy(as high as 90%),and can provide a reliable technical support for low permeability reservoir protection.

关 键 词：量子神经网络 水锁损害 预测 低渗储层 龙西地区 

分 类 号：TE319[石油与天然气工程—油气田开发工程]