沉积微相测井资料神经网络判别方法研究  

The Study on Method of Depositional Microfacies Discrimination with Well-logging Information

作　　者：唐为清[1] 郭荣坤[1] 王忠东 王红[3] 罗安银[4] 毋学平 

机构地区：[1]石油大学,北京100083 [2]辽河石油管理局测井公司,辽宁盘锦124011 [3]四川石油管理局研究院,成都610051 [4]华北石油管理局测井公司,河北任丘062552 [5]华北油田测井公司,河北廊坊102801

出　　处：《沉积学报》2001年第4期581-585,共5页Acta Sedimentologica Sinica

摘　　要：不同的沉积微相可以由不同的相标志组合识别 ,相标志与沉积微相之间的关系可以采用神经网络通过许多基本处理单元间并行的相互作用建立。沉积微相相标志既可以由地质资料的观察、岩芯分析直接获得 。Different depositional microfacies can be identified by the combination of facies signs.The relation between facies signs and depositional microfacies can be established by the parallel reaction of Neural Network basic processing units.Sedimentary microfacies signs can be obtained directly by the observation and analysis of rock core and it can be also got indirectly by well logging information.In this paper, we use modern artificial neural network (ANN) pattern recognition technique to interpret the lithofacies with conventional well logging data and depositional structure with dip well logging data. In determination of lithofacies, eight to ten well logging curve(SP,GR,...) are used. The coincidence rate of lithofacies is eighty to ninety percent in the standard well and seventy to eighty percent in non standard well. ANN was also used to analyze logging facies and depositional environment for single well or multi well on the same sedimentary background. The method of using logging data to automatically identify the carbonate sedimentary microfacies was found out . In the sedimentary microfacies model,24 facies signs that are the combination of lithofacies (category and structure ), sedimentary structure (scoured base, bedding type and the change of laminae ),the direction of paleocurrent, the lithological change, the feature of curve amplitude, the form of curve, the cycle of sedimentation ,the pattern of angle and direction of dip ,pore structure etc. are used. The models offshore deposit and bioherm facies are built up.The method has been used in the field of Si Chuan, Xin Jiang, Liao He and Hua Bei and so on. After testing with some examples, the method is proved to be effective to resolve the problem of petroleum exploration in oil field.

关 键 词：神经网络 沉积微相 相标志 复杂岩性 测井资料 地质资料 岩芯分析 沉积环境 

分 类 号：P588.2[天文地球—岩石学] P512.2[天文地球—地质学]