摘要
利用地质、钻井、测井、三维地震等资料,以民丰地区沙四下亚段为研究对象,提出了膏盐层系砂砾岩体细分层的方法:依据膏盐层期次划分砂砾岩体期次;采用电阻率和声波测井曲线即ΔlgR方法对膏盐层的内部旋回进行划分,进一步细分砂砾岩体亚期;运用加入遗传算法的BP神经网络方法判别砂砾岩体岩性,结合地震奇异性分析,划分扇体内幕。研究表明,加入遗传算法的BP神经网络方法可以对单井的砂砾岩体内幕进行划分,时频谱斜率的奇异性剖面中砂砾岩体包络线内的每个页状体都对应着一组扇体,可以利用其实现砂砾岩体横向的细致划分。民丰地区沙四下亚段的砂砾岩体分为3期、6个亚期、14个砂组,并可进一步细分小层,为勘探开发研究提供分层依据。
According to the geologic, drilling, logging and 3-D seismic data, the method of delamination for glutinite bodies in gypsum-salt bed was advanced as example with the glutinite bodies of the under section of fourth member of Shahejie formation in Minfeng region. The gypsum-salt bed times could be used to carve up glutinite bodies period. The inter cycles of gypsum-salt bed were studied by △lgR method （ the estimate method for organic matter content of mud by resistivity and acoustic logging）, and the sub-period of glutinite bodies can be measured. The lithology was differentiated by BP neural network acceded with genetic algorithm, and the internal fan body was classified, combining with seismic singularity attribute. The results show that the neural network acceded with genetic algorithm could reflect single well lithology sensitively. The seismic singularity attribute section of time-frequency spectrum slope could be used to measure off glutinite bodies transversely. The every lamina in the section was corresponded with one set of fan. With this method, the glutinite bodies of the under section of fourth member of Shahejie formation in Minfeng region were separated into 3 periods, 6 sub-periods, 14 sets and many thin layers. This method afforded delaminafion grounds for the exploration and development.
出处
《中国石油大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2009年第1期1-6,共6页
Journal of China University of Petroleum(Edition of Natural Science)
基金
国家自然科学基金项目(40872094)
山东省博士后创新基金项目(200702040)
关键词
砂砾岩体
膏盐层系
神经网络
奇异性分析
内幕期次
glutinite bodies
gypsum-salt bed
neural network
seismic singularity attribute
internal times
