摘要
针对现存油田开发动态指标预测方法的不足,借助小波神经网络思想、原理,本文构造了小波神经网络油田开发动态指标预测模型,研究了模型的求解,并应用于实际油田开发动态指标的预测中。实例分析表明,提出的小波神经网络的油田开发动态指标的预测方法是正确和可行的。与人工神经网络的预测方法相比,小波神经网络的预测方法不仅能有效地提高预测精度,而且能提高收敛速度,可作为油田开发动态指标预测的替代方法。
In order to overcome the disavantages of the present prediction methods for oilfield development dynamic index,the paper introduces a prediction model of wavelet neural network and its resolution.It has been applied to practice.Case analysis indicates that the method is right and feasible.Comparing with artificial neural network,wavelet neural network prediction method does not only improve the precision of prediction,but also improves rate of convergence.It can be taken as the displacement for oilfield development dynamic indicator prediction.
出处
《石油工业计算机应用》
2008年第2期22-25,共4页
Computer Applications Of Petroleum
基金
四川省教育厅自然科学重点基金项目(编号为:07ZA143)
关键词
油田开发动态指标
预测
小波神经网络(WNN)模型
求解
oilfield development dynamic indicator
prediction
wavelet neural network(WNN)model
resolution
