摘要
随着勘探程度的不断提高 ,未发现的大型构造油气藏已越来越少 ,因而寻找隐蔽的岩性油藏是一项紧迫的任务 ,利用地震资料识别砂岩是完成这一任务的关键。已有的预测砂岩的方法都要求地层厚度在调谐厚度以内 ,对厚层无法准确预测。针对松辽盆地南部砂泥岩薄互层的地质特点 ,从地震数据中提取并优选地震波动力学参数 ,利用神经网络方法学习、训练 ,准确识别预测砂岩的厚度 。
With the increase of survey degree,the big undiscovered structuralreservoirs have become less and less,so it is very important work to search suppressedsandstone reservoirs.The work of using seismic data to identify sandstone can help us to reach the goal of searching for suppressedsandstone reservoirs.The general predicting sandstone methods are usually fit to the layer whose thickness is less than the harmonious',but they are not fit to the thicker thickness.In view of the geologic character of thin interbed in southern Songliao basin,the paper proposed a method that uses seismic dynamicparameters and neural network technique for study, trainingto identify sandstone thickness.The method is suitable even if the thickness is bigger than the harmonious'.
出处
《世界地质》
CAS
CSCD
2002年第2期184-188,共5页
World Geology
关键词
地震波动力学参数
神经网络
油藏
砂岩
seismic dynamic parameter
neural network
reservoir
sandstone
