摘要
针对国外原油气油体积比预测模型在国内一些油田并不适用,在分析前向BP神经网络基本原理的基础上,提出了原油气油体积比新的预测模型。该模型结构为4—10—1的三层BP网络模型,它考虑了压力、温度、地面原油重度和气体的相对密度对气油体积比的影响。利用该模型对大庆油田实测值进行了训练与测试。测试结果表明:利用人工神经网络方法建立的气油体积比预测模型比国外模型精度高,基本合理可靠。
The forecasting models abroad are not well applied in oil field in our country,the new forecasting model of gas-oil ratio has been proposed based on analysis of the basic principle of forward back propagation(BP) neural network.The structure of model is 4—10—1 three-layer BP network.The effect of pressure,temperature,gravitation degree of crude oil in ground and relative density of gas to gas-oil ration are considered.The three BP neural networks are trained and tested by using practical measuring data of Daqing oil field.The result indicates that the new forecasting model is much better in accuracy than those from abroad and it is practical and reliable.
出处
《新疆石油天然气》
CAS
2007年第3期31-33,共3页
Xinjiang Oil & Gas
关键词
气油比
神经网络
预测模型
油田
Gas-oil Ration
Neural Network
Forecasting Model
Oil Field
