论文部分内容阅读
碳酸盐岩裂缝系统中所含流体性质的确定（ 即究竟是含气还是含水或二者均有） 是四川盆地天然气勘探中所面临的关键问题和难点之一。在用川南沈公山构造的三维地震资料提取出多种地震特征参数和综合利用已有的钻井、测井资料的基础上，通过使用具有无教师监督和自聚类特点的自组织映射神经网，提出了一套地震特征参数筛选和气、水识别分类器设计的方法。由于应用该方法能够将沈公山构造上已知井所钻遇裂缝系统中的气、水进行正确区分，因此有可能利用该方法对未钻遇裂缝系统中的流体性质（气或水）作出判断。
The determination of fluid properties in carbonate fracture systems (ie, whether it is gas-bearing or water-bearing, or both) is one of the key issues and difficulties in natural gas exploration in the Sichuan Basin. Based on the three-dimensional seismic data constructed by Shengongshan in southern Sichuan, a variety of seismic parameters were extracted and the existing drilling and well logging data were comprehensively used. By using self-organizing map neural network with no teacher supervision and self-clustering characteristics, A set of seismic feature parameter screening and gas, water recognition classifier design method. Since this method can be used to properly distinguish between gas and water in the fractured system encountered in the structurally known wells in the Shengongshan Mountains, it is possible to use this method to judge the fluid properties (gas or water) in the unbroken fracture system.