摘要
针对川东地区高陡构造下石炭系储层埋藏深、地震反射波信噪比低等待点,提出了一套利用地震资料经人工神经网络训练进行气藏预测的方法,特别是在网络训练前利用模型法对高陡构造和预测点先行归位计算,经实例检验,取得了好的效果。
n artificial neural network natural gas prediction method is put forward using seismic datawith respect to the characteristics of deeper burial reservoir,and lower signal/noise ratio in steep structureregion of Eastern Sichuan Province.Specially the modeling migration technique is adopted for the structureand prediction point position.Good results are obtained using these methods to process two seismic profiles.