井间参数预测的相控神经网络模型

本文提出了一种基于微相研究的神经网络井间参数内插预测新方法。该方法结合油藏微相研究成果，采用井位和微相信息作为神经网络的输入参数，对储层参数进行空间预测。本文以孤岛油田渤21断块油藏为例，利用空间分散井位点的渗透率资料和地区沉积微相信息进行井间渗透率内插预测。结果表明，该方法不仅可以方便地将一些先验的地区知识和专家经验用于井间参数预测之中，而且大大提高了井间参数的预测精度，为油藏建模提供了可靠的基础。

Facies-control neural network model for prediction of inter-well parameters.

The paper presented a new method for prediction of inter well parameters interpolation by using neural network technique based on microfacies study. Combining with studying results of reservoir microfacies,the method uses well position and microfacies information as a input parameters of neural network to make space prediction of reservoir parameters. Taking Bo 21 fault block oil reservoir, Gudao Oilfield as an example,the paper uses permeability data in separate well position and regional depositional microfacies data to make interpolation prediction of inter well permeability. The results show that the method can not only conveniently use prior regional knowledge and expert experiences for prediction of inter well parameters,but also greatly improve the precision of predicted inter well parameters,which provided a reliable basis for reservoir model building.