基于神经网络模型的混合原油黏度预测  

Forecasting Viscosity of Blended Crude Oil Based on Neural Network Model

作　　者：陈坤明[1] 汪玉春[1] 吴华丽[1] 陈进殿[2] 

机构地区：[1]西南石油大学 [2]中国石油天然气股份有限公司规划总院

出　　处：《石油规划设计》2007年第6期33-35,共3页Petroleum Planning & Engineering

摘　　要：采用BP(back propagation)神经网络模拟了混合油配比和温度与黏度之间的映射关系,建立了混合油黏度神经网络预测模型。将鲁-宁输油管道混合油的预测黏度与实际黏度进行了对比,结果表明,利用BP神经网络进行混合油黏度预测是可行的,完全可以满足工程需要的精度要求。The mapped relationship of mixed proportion and temperature with viscosity of the blended oil was simulated by BP-neural-network. The neural network forecasting model for viscosity of the blended oil was presented. The comparison of calculated viscosity with the actual viscosity of the blended oil in Lu-Ning oil pipeline shows that BP-neural-network could be used to predict the viscosity of the blended oil,and the forecasting precision could meet the requirement of the transportation engineering.

关 键 词：神经网络模型 混合原油 黏度计算 预测 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]