摘要
提出一种基于小波神经网络的武器装备研制阶段的费用预测方法.建立小波神经网络预测模型,并推导该模型的预测算法.应用一组美军反舰导弹数据进行仿真预测.结果表明,小波神经网络方法比传统BP网络方法平均预测误差减小了1%,收敛速度加快了3倍.
A new wavelet neural network-based method of forecasting the cost incurred during the development phase of military equipments is proposed.The model of wavelet neural network is constructed and the algorithm implemented.Economical missile data of the American Army is used to do the forecast.It is shown that,compared with the results from traditional BP neural network,the average error rate is reduced by 1% and the convergence speed raised by 3 times.
出处
《北京理工大学学报》
EI
CAS
CSCD
北大核心
2006年第11期991-994,共4页
Transactions of Beijing Institute of Technology
基金
国家部委预研项目(42001020404)
关键词
武器装备
费用预测
小波神经网络
预测模型
预测算法
military equipments
cost forecast
wavelet neural network
forecasting model
forecasting algorithm
