基于GA和ANN的随机规划模型对储层参数的预测  

Based on GA and ANN Random Plan Model for Reservor Parameter Predication

作　　者：朱连章[1] 杨杰[1] 

机构地区：[1]中国石油大学计算机与通信工程学院,东营257061

出　　处：《微计算机应用》2005年第2期218-220,共3页Microcomputer Applications

摘　　要：储层物性参数的分布一直是油藏描述的关键问题和难点。本文采用随机规划与ANN(人工神经网络模型)相结合的建模方法,设计了一种改进的GA(遗传算法)作为随机规划模型的解法,并把这种方法用于储层参数的预测。实验表明该模型具有较强的预测能力和实用性。Distribution of reservoir physical parameter is always the bottleneck problem and difficulty of reservoir description. This paper adopts the modeling method of combination of random plan and ANN (Artificial Neural Networks) , designs a improved GA (Genetic Algorithm) qua the solution of random plan model, and applies the method to predict reservoir parameter. Experiments show the model occupies preferable predication performance and practicability.

关 键 词：油气储层 油藏 随机规划模型 储层参数 预测 GA ANN 人工神经网络 

分 类 号：P618.13[天文地球—矿床学]