摘要
由于构造油气藏相继被发现 ,勘探重点向难度较大和成功率较低的各类隐蔽油气藏转移 ,油气直接检测的研究便成为当今油气勘探领域的热点 ,其方法是利用物探、化探手段检测油气产生的微观效应和异常。作者选择松辽盆地南部让字井地区作为实验区 ,在层序地层、储层及断层封闭性研究及总结前人地表物化探油气预测经验的基础上 ,选取化探酸解烃、ΔC、土壤热释光、测氡及土壤电导率进行了地表实际测量 ,并利用 BP和 SOM两种神经网络方法对物化探综合信息进行了油气预测 ,取得了良好的效果。
As tectonic reservoirs have been discovered one after another,the focal points of exploration have been changed to various subtle traps which will be found with greater difficulty and lower success rate.The study on the direct detection of hydrocarbon then becomes a popular topic in current hydrocarbon exploration domain.The method is to detect the microcosmic effects and anomalies of hydrocarbon by means of geophysical and geochemical prospecting.Taking Rangzijing area in the south of the Songliao Basin as an experimental area,the practical surface measurement is done by the geochemical prospecting of acidolysis hydrocarbon,ΔC,soil thermoluminescence,Rn detection and soil electrical conductivity based on the study of sequence stratigraphy,reservoirs and fault sealing as well as the summary of predecessor′s prediction experience on surface geophysical geochemical prospecting for hydrocarbon.By use of BP and SOM two neural network methods,hydrocarbon prediction is made from integrated information of geophysical geochemical prospecting,and good effects are obtained.
出处
《石油实验地质》
CAS
CSCD
北大核心
2000年第1期71-73,84,共4页
Petroleum Geology & Experiment
关键词
油气预测
神经网络
油气勘探
物理化学勘探
hydrocarbon prediction
BP and SOM neural networks
integrated information
