论文部分内容阅读
在调研了国内外气烟囱研究成果与思路的基础上,通过全新的地震解决方案——OpendTect平台提供的基于属性多层感知器(MLP)的人工神经网络(ANN)的方法预测气烟囱的发生概率体,并利用倾角导向体对算法进行改进,提出倾角控制下的地震气烟囱识别技术,很好地补充并发展了模式识别技术在地震勘探领域的应用。通常该技术一方面可以解释运移通道和浅层气成藏的原因;另一方面,形成的气烟囱地震数据体还可以预测源岩的发育情况;此外,对于判别断层封闭性也非常适用。最后应用这一技术对海拉尔盆地贝尔凹陷一含油区块的运移通道和成藏规律进行研究,分析该研究区的断裂发育特征和运移通道类型,总结了油源、通道、储层和盖层的空间配置关系,并建立相应的成藏模式。
Based on the research results and ideas of gas chimney research at home and abroad, the gas chimney is predicted by the new MLP artificial neural network (ANN) method provided by OpendTect platform. Probability body, and the use of dip-oriented body to improve the algorithm, put forward the gas dip chimney recognition technology under the control of dip angle, well complement and develop the pattern recognition technology in the field of seismic exploration. In general, this technique can explain the reasons for migration channel and shallow gas accumulation. On the other hand, the gas chimney seismic data body can also predict the development of source rocks. In addition, it is also very suitable for distinguishing fault sealing. Finally, this technique was applied to study the migration channel and accumulation pattern of an oil-bearing block in Beier Sag, Hailaer Basin. The fracture development characteristics and types of migration channels in the study area were analyzed. Oil sources, channels, reservoirs and covers Layer space configuration, and establish the corresponding accumulation mode.