【摘要】 气烟囱、断层、盐丘与油气运移直接相关,因而这些地质现象的识别对储层预测至关重要。地质目标体有走向和倾向两个方向属性,如果地震解释不能有效地考虑这两个方向的属性,其结果将与实际有很大的出入。本文应用考虑了地震体倾角和方位角信息的倾角数据体,运用有监督的神经网络分析方法提取能够反映地震异常体的最佳属性,提出一种基于倾角数据体控制下的最优化神经网络地震异常体识别方法。实例研究表明,基于倾角数据体控制下的神经网络分析方法在识别气烟囱方面与常规的多属性神经网络方法相比存在明显的优势,更有利于识别气烟囱等地震异常体。更多还原

【Abstract】 Identifying geological bodies such as gas chimney,fault and salt dome is extremely important for hydrocarbon prediction since they are directly associated with oil & gas migration.These geological bodies have two oriental properties: trend and dip.Interpretation results will be totally different compared with the actual situation if these two oriental properties cannot be correctly considered.In this paper,a method for neural networks gas chimney identification based on steering cube is proposed.With the dip and azimuth information of the steering cube,supervised neural networks analysis extracts best seismic attributes which reflect seismic anomaly.A real data example shows that this method has obvious advantages compared with conventional multi-attributes in chimney identification.更多还原