摘要
储层参数预测在油气勘探与开发中发挥着重要作用。采用神经网络法对储层参数进行预测,建立储层参数(储层砂岩厚度和孔隙度)与目的层地震波属性之间的联系,这种联系是通过人工神经网络来实现的。以地震波属性为输入,储层参数为输出,并以测井资料作为约束条件,指导神经网络学习,进而进行储层参数预测。实际资料处理结果表明,该方法预测的储层孔隙度与砂岩厚度是可靠的。
Reservoir parameter prediction plays an important role in hydrocarbon exploration and production.Neural networks are deployed to predict the reservoir parameters and establish a contact between the reservoir parameters(sandstone thickness and porosity of the reservoir) and the attribute of seismic wave in the target zone.The contact is implemented by neural networks.The seismic wave attribute is used as an input,and the reservoir parameters as output,logging data as constraining condition to guide the neural networks learning,and then predict the reservoir parameters.Data process result indicates the method is reliable for predicting reservoir porosity and sandstone thickness.
出处
《石油天然气学报》
CAS
CSCD
北大核心
2005年第S3期69-70,6,共3页
Journal of Oil and Gas Technology
关键词
储层参数
神经网络
地震波属性
预测
reservoir parameter
neural network
seismic wave attribute
prediction
