用神经网络方法识别碳酸盐岩裂缝系统中的气与水
|
摘要
碳酸盐岩裂缝系统中所含流体性质的确定（ 即究竟是含气还是含水或二者均有） 是四川盆地天然气勘探中所面临的关键问题和难点之一。在用川南沈公山构造的三维地震资料提取出多种地震特征参数和综合利用已有的钻井、测井资料的基础上，通过使用具有无教师监督和自聚类特点的自组织映射神经网，提出了一套地震特征参数筛选和气、水识别分类器设计的方法。由于应用该方法能够将沈公山构造上已知井所钻遇裂缝系统中的气、水进行正确区分，因此有可能利用该方法对未钻遇裂缝系统中的流体性质（气或水）作出判断。
It is one of the key and difficult problems in natural gas exploration in Sichuan Basin to identify the property of the fluid (water or gas or both of them) in carbonate rock fracture systems.On the basis of extracting various seismic feature parameters from the 3 D seismic data of Shengongshan structure in South Sichuan and in combination with the data on drilling and logging,a set of methods selecting seismic feature parameters and designing the classifier of gas water recognition are proposed by use of an unsupervised,self clustering and self organizing mapping neural network.Because the gas and water in the frature systems encountered by the wells drilled in Shengongshan structure can be correctly divided by these methods,it is possible to identify the property of the fluid (gas or water) in the undrilled fracture systems by the same methods.
It is one of the key and difficult problems in natural gas exploration in Sichuan Basin to identify the property of the fluid (water or gas or both of them) in carbonate rock fracture systems.On the basis of extracting various seismic feature parameters from the 3 D seismic data of Shengongshan structure in South Sichuan and in combination with the data on drilling and logging,a set of methods selecting seismic feature parameters and designing the classifier of gas water recognition are proposed by use of an unsupervised,self clustering and self organizing mapping neural network.Because the gas and water in the frature systems encountered by the wells drilled in Shengongshan structure can be correctly divided by these methods,it is possible to identify the property of the fluid (gas or water) in the undrilled fracture systems by the same methods.
引文
1 何樵登主编.地震勘探.北京:地质出版社
2 肖先赐编.现代谱估计.哈尔滨:哈尔滨工业大学出版社,1991
3 何振亚.数字信号处理的理论与应用.北京:人民邮电出版社,1983
4 Marple L.A new autoregressive Spectrum analysis.IEEETransactions on Acoustics,Speech,and Signal Processing,1980 ;28(4)
5 Kohonen T.The selforganizing map.Proceedings of TheIEEE,1990;78(9) :1464 ～1480
6 焦李成.神经网络系统理论.西安:西安电子科技大学出版社,1990
7 焦李成编.神经网络计算.西安:西安电子科技大学出版社,1993
8 李亚林.自组织映射神经网油气识别系统研究及其在油气预测中的应用.西南石油学院硕士学位论文,1996
2 肖先赐编.现代谱估计.哈尔滨:哈尔滨工业大学出版社,1991
3 何振亚.数字信号处理的理论与应用.北京:人民邮电出版社,1983
4 Marple L.A new autoregressive Spectrum analysis.IEEETransactions on Acoustics,Speech,and Signal Processing,1980 ;28(4)
5 Kohonen T.The selforganizing map.Proceedings of TheIEEE,1990;78(9) :1464 ～1480
6 焦李成.神经网络系统理论.西安:西安电子科技大学出版社,1990
7 焦李成编.神经网络计算.西安:西安电子科技大学出版社,1993
8 李亚林.自组织映射神经网油气识别系统研究及其在油气预测中的应用.西南石油学院硕士学位论文,1996
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心