利用交替投影神经网络进行地震弱信息分离
|
摘要
当前岩性油气藏勘探面对的储集体大多表现为地震微弱同相轴,在地震剖面上不容易识别,为此,将交替投影神经网络算法引入到地震资料处理中。在分析交替投影神经网络基本理论方法的基础上,研究利用该方法进行地震弱反射信号的分离,并设计实际地质模型进行算法验证,最后对东营凹陷南斜坡高8地区实际地震资料进行计算处理,结果显示,常规地震剖面上的薄砂体地震弱反射信息得到明显改善。理论模型和实际资料处理结果表明,交替投影神经网络算法是进行地震弱反射信息分离的一种有效方法。
At present,the reservoir rocks in the lithologic hydrocarbon reservoirs are mostly characterized by weak seismic event,not easy to be identified on the seismic section. Therefore,the algorithm of alternating project neural network was introduced to seismic data processing. Based on the principles and methods of the alternating project neural network,the algorithm was studied and applied for separating weak reflected signals so as to design actual geological model for validating the algorithm. The results of computation process on actual seismic data from Gao 8 in the southern slope of Dongying Depression show that the weak reflected signals of thin sand body on conventional seismic sections have been improved distinctly. And theoretical model and the results of actual data process also show that alternating project neural network algorithm is available for separating weak reflected signals.
At present,the reservoir rocks in the lithologic hydrocarbon reservoirs are mostly characterized by weak seismic event,not easy to be identified on the seismic section. Therefore,the algorithm of alternating project neural network was introduced to seismic data processing. Based on the principles and methods of the alternating project neural network,the algorithm was studied and applied for separating weak reflected signals so as to design actual geological model for validating the algorithm. The results of computation process on actual seismic data from Gao 8 in the southern slope of Dongying Depression show that the weak reflected signals of thin sand body on conventional seismic sections have been improved distinctly. And theoretical model and the results of actual data process also show that alternating project neural network algorithm is available for separating weak reflected signals.
引文
[1]匡建超,曾剑毅,王众.模糊优选神经网络储层识别技术在长庆中部气田马五1段的应用[J].油气地质与采收率,2008,15(5):5-7,12.
[2]汪嘉月,奥立德,屈红,等.运用神经网络法研究微裂缝的分布规律——以苏北盆地高邮凹陷CA油田为例[J].石油实验地质,2006,28(4):395-398.
[3]李海燕,彭仕宓.应用遗传神经网络研究低渗透储层成岩储集相——以胜利渤南油田三区沙河街组为例[J].石油与天然气地质,2006,27(1):111-117.
[4]刘勇健,沈军,刘义建,等.基于人工神经网络的岩石含油气性评价方法[J].石油实验地质,2000,22(3):276-279.
[5]符翔,韩国庆,安永生.应用人工神经网络优化五点法水平井井网[J].油气地质与采收率,2007,14(2):93-95.
[6]邴绍献,王华,李建丽,等.改进结构的小波神经网络在油田开发指标预测中的应用[J].油气地质与采收率,2009,16(3):92-94.
[7]Marks R J II,Oh S,Atlas L E.Alternating projection neural art-works[J].IEEE Transactions on Circuits and Systems,1989,36(6):846-857.
[8]Youla D C.Generalized image restoration by the method of alterna-ting orthogonal projections[J].IEEE Transactions on Circuits and Systems,1978,25(2):694-702.
[9]Tank D W,Hopfield J I.Simple“neuron”optimization networks:An A/D converter,signal decision circuit,and a linear program-ming circuit[J].IEEE Transactions on Circuits and Systems,1986,33(1):533-541.
[10]王金根,陈世福,龚沈光,等.一种基于扩展交替投影神经网络的弱信号分离方法研究[J].计算机科学,2003,30(10):64-66.
[11]王金根,林春生,龚沈光.基于交替投影神经网络的带限信号外推算法[J].电子学报,2000,28(10):52-55.
[12]张胜伟,杨宗凯,邹理和.基于投影原理的双向联想神经网络[J].电子学报,1991,19(3):23-30.
[13]焦李成.神经网络系统理论[M].西安:西安电子科技大学出版社,1996:45-51.
[14]庞亚红,毛幼菊.一种基于凸集投影(POCS)的数字图像超分辨率重建算法[J].计算机工程与应用,2005,41(4):69-71.
[15]王金根,林春生,龚沈光.复交替投影神经网络[J].数据采集与处理,2001,16(1):18-23.
[16]Geman S,Geman D.Stochastic relaxation,Gibb′s distributions,and the Bayesian restoration of images[J].IEEE Transactions on Pat-tern Analysis and Machine Intelligence,1984,6(6):721-741.
[17]Atlas L E.Auditory coding in higher centers of the CNS[J].IEEE Engineering in Medicine and Biology Magazine,1987,6(2):29-32.
[18]Abu-Mostafa Y S,Jaques St Jeannine-Marie.Information capac-ity of the Hopfield model[J].IEEE Transactions on Information Theory,1985,3l(4):461-467.
[19]Rumelhart D E,Hinton G E,Willians R J.Learning representations by back-propagation errors[J].Nature,1986,323(6088):533-536.
[20]Stanfill C,Waltz D.Toward memory-based reasoning[J].Com-munications of the ACM,1986,29(12):1213-1228.
[2]汪嘉月,奥立德,屈红,等.运用神经网络法研究微裂缝的分布规律——以苏北盆地高邮凹陷CA油田为例[J].石油实验地质,2006,28(4):395-398.
[3]李海燕,彭仕宓.应用遗传神经网络研究低渗透储层成岩储集相——以胜利渤南油田三区沙河街组为例[J].石油与天然气地质,2006,27(1):111-117.
[4]刘勇健,沈军,刘义建,等.基于人工神经网络的岩石含油气性评价方法[J].石油实验地质,2000,22(3):276-279.
[5]符翔,韩国庆,安永生.应用人工神经网络优化五点法水平井井网[J].油气地质与采收率,2007,14(2):93-95.
[6]邴绍献,王华,李建丽,等.改进结构的小波神经网络在油田开发指标预测中的应用[J].油气地质与采收率,2009,16(3):92-94.
[7]Marks R J II,Oh S,Atlas L E.Alternating projection neural art-works[J].IEEE Transactions on Circuits and Systems,1989,36(6):846-857.
[8]Youla D C.Generalized image restoration by the method of alterna-ting orthogonal projections[J].IEEE Transactions on Circuits and Systems,1978,25(2):694-702.
[9]Tank D W,Hopfield J I.Simple“neuron”optimization networks:An A/D converter,signal decision circuit,and a linear program-ming circuit[J].IEEE Transactions on Circuits and Systems,1986,33(1):533-541.
[10]王金根,陈世福,龚沈光,等.一种基于扩展交替投影神经网络的弱信号分离方法研究[J].计算机科学,2003,30(10):64-66.
[11]王金根,林春生,龚沈光.基于交替投影神经网络的带限信号外推算法[J].电子学报,2000,28(10):52-55.
[12]张胜伟,杨宗凯,邹理和.基于投影原理的双向联想神经网络[J].电子学报,1991,19(3):23-30.
[13]焦李成.神经网络系统理论[M].西安:西安电子科技大学出版社,1996:45-51.
[14]庞亚红,毛幼菊.一种基于凸集投影(POCS)的数字图像超分辨率重建算法[J].计算机工程与应用,2005,41(4):69-71.
[15]王金根,林春生,龚沈光.复交替投影神经网络[J].数据采集与处理,2001,16(1):18-23.
[16]Geman S,Geman D.Stochastic relaxation,Gibb′s distributions,and the Bayesian restoration of images[J].IEEE Transactions on Pat-tern Analysis and Machine Intelligence,1984,6(6):721-741.
[17]Atlas L E.Auditory coding in higher centers of the CNS[J].IEEE Engineering in Medicine and Biology Magazine,1987,6(2):29-32.
[18]Abu-Mostafa Y S,Jaques St Jeannine-Marie.Information capac-ity of the Hopfield model[J].IEEE Transactions on Information Theory,1985,3l(4):461-467.
[19]Rumelhart D E,Hinton G E,Willians R J.Learning representations by back-propagation errors[J].Nature,1986,323(6088):533-536.
[20]Stanfill C,Waltz D.Toward memory-based reasoning[J].Com-munications of the ACM,1986,29(12):1213-1228.
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