论文部分内容阅读
本文介绍了人工神经网络在油藏非均质性描述中新的应用方法及研究开发神经网络的实例。在非均质性极高的地层中，孔隙度、渗透率和流体饱和度等不同的油藏性质能用可信、可靠的地球物理测井曲线推演的信息进行精确的预测，但预测的方法要以人工神经网络（三层前馈、后传播）的智能性与自适应性认识能力对基础。它能大大缩短孔隙度、渗透率和流体饱和度数据的高费用参数采集过程（如试井和广义地层取芯等）。
This paper introduces the new application of artificial neural network in reservoir heterogeneity description and examples of neural network. In very heterogeneous formations, different reservoir properties such as porosity, permeability and fluid saturation can be accurately predicted using information derived from credible and reliable geophysical well logs, but the predictive method To artificial neural network (three-feed-forward, after the spread of intelligence and adaptive understanding of the basis. It greatly reduces the costly parameter acquisition process for porosity, permeability, and fluid saturation data (eg, well testing and generalized core formation, etc.).