基于灰关联和遗传算法的地震属性优化方法
摘要
利用地震属性进行储层物性参数(孔隙度)的预测研究,其精度和准确性一直是油气田勘探开发期间所关注的问题,而属性优化方法是其中的关键环节。本文从优化方法入手,将灰关联、遗传算法和BP网络有机结合,以提高储层物性参数(孔隙度)的预测能力。首先依据反演成果开展精细储层解释,并进行沿层相关属性提取和灰关联分析,寻找较敏感的属性;然后利用遗传算法和BP网络进行属性组合优化,获取最优属性组合,最终达到预测储层物性参数(孔隙度)的目的。应用实例表明,此法的储层预测精度较高。
When applying seismic attributes in prediction studies of reservoir physical property (porosity),its precision and accuracy are always the main concern in the oil-field exploration and development,and the attribute optimization method is the focal point.Based on optimization method the rational integration of grey relational analysis,genetic algorithm and BP network could raise the prediction ability of reservoir physical property (porosity).The work can be done as below,at first using inversion results to conduct detailed reservoir interpretation,extracting related attributes along the layers and conducting grey relational analysis and looking for sensitive attributes,then utilizing genetic algorithm and BP network to optimize the attribute combination,and obtaining optimum attribute combination,at last the prediction of reservoir physical property (porosity) was realized.The application results show that the method is a precise in reservoir prediction.
When applying seismic attributes in prediction studies of reservoir physical property (porosity),its precision and accuracy are always the main concern in the oil-field exploration and development,and the attribute optimization method is the focal point.Based on optimization method the rational integration of grey relational analysis,genetic algorithm and BP network could raise the prediction ability of reservoir physical property (porosity).The work can be done as below,at first using inversion results to conduct detailed reservoir interpretation,extracting related attributes along the layers and conducting grey relational analysis and looking for sensitive attributes,then utilizing genetic algorithm and BP network to optimize the attribute combination,and obtaining optimum attribute combination,at last the prediction of reservoir physical property (porosity) was realized.The application results show that the method is a precise in reservoir prediction.
引文
[1]Balch A H.Color sonograms-a new di mension in seis-mic data interpretation.Geophysics,1971,36(1):1074～1098
[2]Taner MTet al.Complex seismic trace analysis.Geo-physics,1979,44:1041～1063
[3]Bahorich Mand Farmer S.3-Dseismic discontinuity for faults and stratigraphic features:The Coherence Cube.SEG Technical Program Expanded Abstracts,1995,14:93～96
[4]Taner M T.Attributes revisited.Rock Solid Images,1992
[5]Chen Q and Sidney S.Advances in seismic attribute technology.SEG Technical Program Expanded Ab-stracts,1997,16:730～733
[6]Chen Qand Sidney S.Seismic attribute technology for reservoir forecasting and monitoring.The Leading Edge,1997,16(5):445～450
[7]孙即祥.现代模式识别.湖南长沙:国防科技大学出版社,2002
[8]赵加凡,陈小宏.基于主成分分析与KL变换的双重属性优化方法.物探与化探,2005,29(3):253～255
[9]童其慧.主成分分析方法在指标综合评价中的应用.北京理工大学学报:社会科学版,2002,4(l):59～61.
[10]陈遵德.储层地震属性优化方法.北京:石油工业出版社,1998:25～46
[11]邓聚龙.灰色系统理论教程.湖北武汉:华中理工大学出版社,1990:5～264
[12]季敏,王尚旭,陈双全.灰色理论在地球物理勘探开发中的应用综述.地球物理学进展,2005,20(4):1164～1170
[13]赵加凡,陈小宏,张勤.灰关联分析在储层评价中的应用.勘探地球物理进展,2003,26(4):281-286
[14]李谋杰,毛宁波,张敏知.基于灰色系统理论的储层参数预测方法与应用.石油天然气学报(江汉石油学院学报),2006,28(3):64～67
[15]Goldberg D E.Genetic Algorithms in Search,Opti-mization and Machine Learning.Addison-Wesley Publishing Company,1989
[16]Sen M K and Stoffa P L.Rapid sampling of model space using genetic algorithms:Examples from seis-mic waveforminversion.Geophys J Internat,1992,108:281～292
[17]陈遵德,朱广生等.用遗传算法选择储层预测中的地震特征.中国地球物理学会年刊.北京:石油工业出版社,1995
[18]王永刚.地震属性的GA-BP优化方法.石油地球物理勘探,2002,37(6):606～611
[19]曹思远.储层预测中BP神经网络的应用.地球物理学进展,2002,17(1):84～90
[20]黄捍东,罗群,付艳,王保华.地震相控非线性随机反演研究与应用.石油地球物理勘探,2007,42(6):694～698
[21]DengJulong.Control problemof greysystems.System&Control Letters,1982,3
[22]Rumellhart D,Hinton G,Williams R.Learning rep-resentation by back-propagating errors.Nature,1986,323:533～536
[2]Taner MTet al.Complex seismic trace analysis.Geo-physics,1979,44:1041～1063
[3]Bahorich Mand Farmer S.3-Dseismic discontinuity for faults and stratigraphic features:The Coherence Cube.SEG Technical Program Expanded Abstracts,1995,14:93～96
[4]Taner M T.Attributes revisited.Rock Solid Images,1992
[5]Chen Q and Sidney S.Advances in seismic attribute technology.SEG Technical Program Expanded Ab-stracts,1997,16:730～733
[6]Chen Qand Sidney S.Seismic attribute technology for reservoir forecasting and monitoring.The Leading Edge,1997,16(5):445～450
[7]孙即祥.现代模式识别.湖南长沙:国防科技大学出版社,2002
[8]赵加凡,陈小宏.基于主成分分析与KL变换的双重属性优化方法.物探与化探,2005,29(3):253～255
[9]童其慧.主成分分析方法在指标综合评价中的应用.北京理工大学学报:社会科学版,2002,4(l):59～61.
[10]陈遵德.储层地震属性优化方法.北京:石油工业出版社,1998:25～46
[11]邓聚龙.灰色系统理论教程.湖北武汉:华中理工大学出版社,1990:5～264
[12]季敏,王尚旭,陈双全.灰色理论在地球物理勘探开发中的应用综述.地球物理学进展,2005,20(4):1164～1170
[13]赵加凡,陈小宏,张勤.灰关联分析在储层评价中的应用.勘探地球物理进展,2003,26(4):281-286
[14]李谋杰,毛宁波,张敏知.基于灰色系统理论的储层参数预测方法与应用.石油天然气学报(江汉石油学院学报),2006,28(3):64～67
[15]Goldberg D E.Genetic Algorithms in Search,Opti-mization and Machine Learning.Addison-Wesley Publishing Company,1989
[16]Sen M K and Stoffa P L.Rapid sampling of model space using genetic algorithms:Examples from seis-mic waveforminversion.Geophys J Internat,1992,108:281～292
[17]陈遵德,朱广生等.用遗传算法选择储层预测中的地震特征.中国地球物理学会年刊.北京:石油工业出版社,1995
[18]王永刚.地震属性的GA-BP优化方法.石油地球物理勘探,2002,37(6):606～611
[19]曹思远.储层预测中BP神经网络的应用.地球物理学进展,2002,17(1):84～90
[20]黄捍东,罗群,付艳,王保华.地震相控非线性随机反演研究与应用.石油地球物理勘探,2007,42(6):694～698
[21]DengJulong.Control problemof greysystems.System&Control Letters,1982,3
[22]Rumellhart D,Hinton G,Williams R.Learning rep-resentation by back-propagating errors.Nature,1986,323:533～536
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