小波神经网络在原油含水率预测中的应用研究  

Application Research of Wavelet Neural Network in Prediction of Water Cut of Crude Oil

作　　者：王乃民[1] 刘润华[1] 肖广云[1] 

机构地区：[1]中国石油大学信息与控制工程学院,山东东营257061

出　　处：《化工自动化及仪表》2008年第3期51-53,57,共4页Control and Instruments in Chemical Industry

摘　　要：通过对原油含水率测量方法,以及原油含水率与影响其因素之间存在的非线性映射关系的研究,提出基于小波分解与神经网络相结合的小波神经网络原油含水率预测模型,给出具体的网络学习算法,并结合算法对原油含水率进行预测。实例分析表明,小波神经网络模型比传统的BP神经网络模型收敛速度快、预测精度高,且具有较强的学习能力和推广能力。According to the research on the measuring method of water cut of crude oil, as well as the nonlinear map- ping relationship between the water cut of crude oil and the affecting factor, this paper brought up a forecasting model of water cut of crude oil based on wavelet decomposition and neural network. Besides, detailed learning algorithm was proposed and it was used in prediction. The case analysis shows that, the utilization of neural network generates more fast convergence rate, more precise forecast than traditional BP neural network model, and it presents good abilities of learning and dissemination

关 键 词：小波神经网络 原油含水率 预测 非线性 BP神经网络 

分 类 号：TE866[石油与天然气工程—油气储运工程]