基于粗集理论的多极阵列声波测井数据处理方法研究
|
摘要
声波测井是近年来发展较快的一种测井方法。将声波测井数据与地质勘探的观测资料结合起来,在判断岩性,识别压力异常层位,探测和评价裂缝,判断储集层中流体的性质方面,有着非常良好的发展前景。
本文结合电子科技大学承担的中海油企业发展基金项目“数字阵列声波仪器EDAT项目研究”这一课题展开研究,阐述了多极阵列声波测井原理及基本方法,分析了多极阵列声波测井与地质勘探的观测资料的数据特征,研究了基于粗集理论的多极阵列声波测井数据的重要特征的新的提取算法。并结合模糊聚类理论,提出了基于粗集理论的多极阵列声波测井数据处理的系统决策方法。本文为解决多极子阵列声波测井仪测量数据的后续处理,判断岩性,识别压力异常层位,探测和评价裂缝,判断储集层中流体的性质等方面提供了理论依据,奠定了实现方案的基础。
本文的主要工作如下:
1、根据岩石的声学特性,介绍了声波速度测井、声波幅度测井、长源距声波全波列测井的基本原理,分析了多极阵列声波测井系统的实现及其测井工作中的几个关键技术。
2、根据粗集理论的方法,介绍了多极阵列声波测井的信息系统的基本概念,研究了知识及系统的简化方法,以及系统参数的重要性评价算法的改进。并结合一些分析实例,给出了基于粗集理论的多极阵列声波测井数据的重要特征的分析和提取方法。
3、根据粗集理论,结合模糊聚类理论,研究了协调数据的决策规则简化算法及信息覆盖率在评价方面的改进,并结合一些分析实例,给出了基于粗集理论的多极阵列声波测井数据的最简化系统处理和提取决策规则的分析方法。
Sonic logging is one of logging methods developed in recent years. Sonic logging has a very good foreground when it is combined with geologic reconnoissance, especially in respects of lithology judgment, abnormal pressure layer recognition, crack estimation and judgment of liquid characteristic in store layer.
The topic of this thesis is based on the project of "an instrument product of digital array sonic logging" taken by university of electronic science and technology of China, which is supported by sea oil company of China. In the thesis, an expatiation of the basic theory and technology of muti-pole sonic logging is given firstly. Then a kind of selection and reduction of significant information features of multi-pole sonic logging based on rough sets is presented. A new approach of data processing and recognition based on fuzzy clustering of rough sets is also put forword as theoretical foundation to process the continued data of muti-pole sonic logging, lithology judgment abnormal pressure layer recognition, crack estimation and judgment of liquid character in store layer.
The main works in the thesis are as follows:
1. Based on rocky acoustic characteristic, a theory of sonic speed logging, sonic amplitude logging, long-distance sonic logging is intrduced and some main techniques of muti-pole sonic logging are analysed.
2. By means of rough sets, a basic conceptioal introduce of information system of muti-pole sonic logging is interpreted, then a kind of improvement algorithm of system parameter's importance is put forward. The reduction of knowledge and presentation of system is investigated by some examples. Thus a new approach of reduction of attributes of multi-pole sonic logging data based on rough sets is presented in the thesis finally.
3. Combined with fuzzy cluster techniques, a kind of improvement algorithm of decision-making rules is investigated and a new method of system reduction by information coverage is proposed through some examples. At last, a new system of decision-making rules in multi-pole sonic logging data mining is set up.
本文结合电子科技大学承担的中海油企业发展基金项目“数字阵列声波仪器EDAT项目研究”这一课题展开研究,阐述了多极阵列声波测井原理及基本方法,分析了多极阵列声波测井与地质勘探的观测资料的数据特征,研究了基于粗集理论的多极阵列声波测井数据的重要特征的新的提取算法。并结合模糊聚类理论,提出了基于粗集理论的多极阵列声波测井数据处理的系统决策方法。本文为解决多极子阵列声波测井仪测量数据的后续处理,判断岩性,识别压力异常层位,探测和评价裂缝,判断储集层中流体的性质等方面提供了理论依据,奠定了实现方案的基础。
本文的主要工作如下:
1、根据岩石的声学特性,介绍了声波速度测井、声波幅度测井、长源距声波全波列测井的基本原理,分析了多极阵列声波测井系统的实现及其测井工作中的几个关键技术。
2、根据粗集理论的方法,介绍了多极阵列声波测井的信息系统的基本概念,研究了知识及系统的简化方法,以及系统参数的重要性评价算法的改进。并结合一些分析实例,给出了基于粗集理论的多极阵列声波测井数据的重要特征的分析和提取方法。
3、根据粗集理论,结合模糊聚类理论,研究了协调数据的决策规则简化算法及信息覆盖率在评价方面的改进,并结合一些分析实例,给出了基于粗集理论的多极阵列声波测井数据的最简化系统处理和提取决策规则的分析方法。
Sonic logging is one of logging methods developed in recent years. Sonic logging has a very good foreground when it is combined with geologic reconnoissance, especially in respects of lithology judgment, abnormal pressure layer recognition, crack estimation and judgment of liquid characteristic in store layer.
The topic of this thesis is based on the project of "an instrument product of digital array sonic logging" taken by university of electronic science and technology of China, which is supported by sea oil company of China. In the thesis, an expatiation of the basic theory and technology of muti-pole sonic logging is given firstly. Then a kind of selection and reduction of significant information features of multi-pole sonic logging based on rough sets is presented. A new approach of data processing and recognition based on fuzzy clustering of rough sets is also put forword as theoretical foundation to process the continued data of muti-pole sonic logging, lithology judgment abnormal pressure layer recognition, crack estimation and judgment of liquid character in store layer.
The main works in the thesis are as follows:
1. Based on rocky acoustic characteristic, a theory of sonic speed logging, sonic amplitude logging, long-distance sonic logging is intrduced and some main techniques of muti-pole sonic logging are analysed.
2. By means of rough sets, a basic conceptioal introduce of information system of muti-pole sonic logging is interpreted, then a kind of improvement algorithm of system parameter's importance is put forward. The reduction of knowledge and presentation of system is investigated by some examples. Thus a new approach of reduction of attributes of multi-pole sonic logging data based on rough sets is presented in the thesis finally.
3. Combined with fuzzy cluster techniques, a kind of improvement algorithm of decision-making rules is investigated and a new method of system reduction by information coverage is proposed through some examples. At last, a new system of decision-making rules in multi-pole sonic logging data mining is set up.
引文
[1]张松扬,国外声波测井技术新发展.石油物探译丛,1998,(5):1-10
[2]张辛耕,随钻测井技术进展和发展趋势.测井技术,2006,30(1):10-15
[3]谭廷栋等译,测井解释基础与数据采集,石油工业出版社,1992,1:1-320
[4]刘文龙,相控声波测井仪器设计及资料处理方法.石油仪器,2006,20(3):14-17
[5]沈建国,高分辨率声波测井面临的问题及其对策.石油地球物理勘探,2006,41(1):111-115
[6]魏岩峻,王自强,刘海涛,等,一种新型的声波测井系统及其数据处理.人民长江,33(6),2002:28-30
[7]D.W.希尔契,现代测井解释.石油工业出版社,1989,4:226-235
[8]陶果,邹辉,现代阵列声波测井数据处理和解释方法研究.石油勘探与开发,2000,27(2):76-78
[9]Chen S T.,Shear Wave Logging with dipole Sources.Geophysics,1988,(53):659-667
[10]李长文,强毓明,李国利,等,现代声波测井技术发展的若干特点.测井技术,2006,30(2):101-104
[11]末延杰,何英伟,石颖,等,慢度时间相关法与遗传算法结合提取阵列声波时差.测井技术,2006,4,30(2):122-125
[12]多极子阵列声波测井仪技术报告,中海油田服务股份有限公司.2003,11:3-26
[13]陈必孝,张筠,声波全波列测井资料分析处理技术及应用.测井技术,2002,26(5):369-372
[14]Z.Pawlak,Rough Sets,Theoretical Aspects of Reasoning about Data.Nowowiejska,Warsaw,Poland,1990:15-19
[15]曾黄麟,智能计算.重庆:重庆大学出版社,2004,6:31-64
[16]易淑昌,基于粗集理论的数据挖掘技术识别储层岩性.国外测井技术,2003,18(5):12-15
[17]王国胤,决策表核属性的计算方法.计算机学报,2003,26(5):611-615
[18]郑征,王国胤,吴渝,基于合并规则的决策表求核方法.计算机科学,2002,29(9):204-207
[19]夏克文,沈钧毅,李昌彪,一种基于计算智能的油气层识别方法.计算机工程与应用,2005(5):190-192
[20]Zenghuanglin,Lanhengyou,Zengxiaohui,A New Method of Selection and Reduction of System Feature in Pattern Recognition based on Rough Sets.Journal of Communication and Computer,2006,Aug.,3(8):25-28
[21]杨成福,舒兰,基于属性重要性的决策表属性约简算法.计算机技术与发展,2006,16(11):62-66
[22]曾晓辉,练艺,课堂教学质量的重要因素分析.四川师范大学学报,2006,29(1):231-234
[23]李昌彪,宋建平,夏克文,属性约简在测井信息优化中的应用.微电子学与计算机,2003,8:130-133
[24]曾晓辉,师奕兵,练艺,一种多极阵列声波测井重要特征分析方法.四川理工学院学报,2007,20(1):48-51
[25]Zengxiaohui,Shiyibing,Lianyi,An Interpretation of Muti-pole Sonic Logging Data Mining Based on Rough Sets.Communication and Computer,2007,4(1):8-10
[26]李昌彪,夏克文,宋建平,等,一种基于属性重要性的粗糙RBF神经网络.控制与决策,2006,21(7):821-824
[27]郭海湘,诸克军,李玥,等,软计算与硬计算融合的中国石油需求预测.中国地质大学学报,2007,6(6):55-67
[28]夏克文,李昌彪,刘宏杰,等,用于油水层识别的一种简化的神经计算方法.大庆石油地质与开发,2006,25(3):56-59
[29]Zengxiaohui,Shiyibing,Lianyi,An Interpretation of Muti-pole Sonic Logging Data Mining Based on Rough Sets.Communication and Computer,2007,4(1):8-10
[30]王士同,模糊系统、模糊神经网络及应用程序设计.上海科学技术文献出版社,1997,9:1-20
[31]李汉林,连承波,马士坤,等,基于气测资料的储层含油气性识别方法.中国石油大学学报,2006.30(4):21-23
[32]曾黄麟,师奕兵,曾晓辉,一种新的基于粗集编码的模糊聚类数据处理方法.电子测量与仪器学报,2007.21(3):39-42
[33]Zenghuanglin,Yuanzengren,Zengxiaohui,Redundant Data Processing Based on Rough-Fuzzy Approach.Rough Sets and Knowledge Technology,Springer-Verlag Berlin Heidelberg,Germany,2006:156-161
[34]夏克文,宋建平,李昌彪,基于粗集和神经网络在石油测井数据挖掘方法.信息与控制,2003,8,32(4):300-303
[35]李昌彪,宋建平,夏克文,利用RBF神经网络自适应调整算法预测储层产能.石油地球物理勘探,2006,2,41(1):53-57
[36]罗利,姚声贤,任兴国,等,神经网络及模式识别技术在测井解释中的应用.测井技术,2002,26(5):364-368
[2]张辛耕,随钻测井技术进展和发展趋势.测井技术,2006,30(1):10-15
[3]谭廷栋等译,测井解释基础与数据采集,石油工业出版社,1992,1:1-320
[4]刘文龙,相控声波测井仪器设计及资料处理方法.石油仪器,2006,20(3):14-17
[5]沈建国,高分辨率声波测井面临的问题及其对策.石油地球物理勘探,2006,41(1):111-115
[6]魏岩峻,王自强,刘海涛,等,一种新型的声波测井系统及其数据处理.人民长江,33(6),2002:28-30
[7]D.W.希尔契,现代测井解释.石油工业出版社,1989,4:226-235
[8]陶果,邹辉,现代阵列声波测井数据处理和解释方法研究.石油勘探与开发,2000,27(2):76-78
[9]Chen S T.,Shear Wave Logging with dipole Sources.Geophysics,1988,(53):659-667
[10]李长文,强毓明,李国利,等,现代声波测井技术发展的若干特点.测井技术,2006,30(2):101-104
[11]末延杰,何英伟,石颖,等,慢度时间相关法与遗传算法结合提取阵列声波时差.测井技术,2006,4,30(2):122-125
[12]多极子阵列声波测井仪技术报告,中海油田服务股份有限公司.2003,11:3-26
[13]陈必孝,张筠,声波全波列测井资料分析处理技术及应用.测井技术,2002,26(5):369-372
[14]Z.Pawlak,Rough Sets,Theoretical Aspects of Reasoning about Data.Nowowiejska,Warsaw,Poland,1990:15-19
[15]曾黄麟,智能计算.重庆:重庆大学出版社,2004,6:31-64
[16]易淑昌,基于粗集理论的数据挖掘技术识别储层岩性.国外测井技术,2003,18(5):12-15
[17]王国胤,决策表核属性的计算方法.计算机学报,2003,26(5):611-615
[18]郑征,王国胤,吴渝,基于合并规则的决策表求核方法.计算机科学,2002,29(9):204-207
[19]夏克文,沈钧毅,李昌彪,一种基于计算智能的油气层识别方法.计算机工程与应用,2005(5):190-192
[20]Zenghuanglin,Lanhengyou,Zengxiaohui,A New Method of Selection and Reduction of System Feature in Pattern Recognition based on Rough Sets.Journal of Communication and Computer,2006,Aug.,3(8):25-28
[21]杨成福,舒兰,基于属性重要性的决策表属性约简算法.计算机技术与发展,2006,16(11):62-66
[22]曾晓辉,练艺,课堂教学质量的重要因素分析.四川师范大学学报,2006,29(1):231-234
[23]李昌彪,宋建平,夏克文,属性约简在测井信息优化中的应用.微电子学与计算机,2003,8:130-133
[24]曾晓辉,师奕兵,练艺,一种多极阵列声波测井重要特征分析方法.四川理工学院学报,2007,20(1):48-51
[25]Zengxiaohui,Shiyibing,Lianyi,An Interpretation of Muti-pole Sonic Logging Data Mining Based on Rough Sets.Communication and Computer,2007,4(1):8-10
[26]李昌彪,夏克文,宋建平,等,一种基于属性重要性的粗糙RBF神经网络.控制与决策,2006,21(7):821-824
[27]郭海湘,诸克军,李玥,等,软计算与硬计算融合的中国石油需求预测.中国地质大学学报,2007,6(6):55-67
[28]夏克文,李昌彪,刘宏杰,等,用于油水层识别的一种简化的神经计算方法.大庆石油地质与开发,2006,25(3):56-59
[29]Zengxiaohui,Shiyibing,Lianyi,An Interpretation of Muti-pole Sonic Logging Data Mining Based on Rough Sets.Communication and Computer,2007,4(1):8-10
[30]王士同,模糊系统、模糊神经网络及应用程序设计.上海科学技术文献出版社,1997,9:1-20
[31]李汉林,连承波,马士坤,等,基于气测资料的储层含油气性识别方法.中国石油大学学报,2006.30(4):21-23
[32]曾黄麟,师奕兵,曾晓辉,一种新的基于粗集编码的模糊聚类数据处理方法.电子测量与仪器学报,2007.21(3):39-42
[33]Zenghuanglin,Yuanzengren,Zengxiaohui,Redundant Data Processing Based on Rough-Fuzzy Approach.Rough Sets and Knowledge Technology,Springer-Verlag Berlin Heidelberg,Germany,2006:156-161
[34]夏克文,宋建平,李昌彪,基于粗集和神经网络在石油测井数据挖掘方法.信息与控制,2003,8,32(4):300-303
[35]李昌彪,宋建平,夏克文,利用RBF神经网络自适应调整算法预测储层产能.石油地球物理勘探,2006,2,41(1):53-57
[36]罗利,姚声贤,任兴国,等,神经网络及模式识别技术在测井解释中的应用.测井技术,2002,26(5):364-368