摘要
针对现代战争条件下装备保障资源需求变化快,保障资源预测困难的问题,首先分析了影响装备保障资源需求的因素,根据实际情况选取了平均维修间隔时间(MTBM)、平均修复时间(MTBR)等8项影响装备保障资源需求的关键指标,然后将基于遗传算法(GA)优化的反向传播(BP)神经网络应用于保障资源需求预测中,构建了基于遗传神经网络的需求预测模型,最后利用1980年～2010年实际保障资源需求数据对模型进行了验证。验证结果表明,基于GA优化的BP神经网络预测模型有较快的收敛速度、较强的适应性和较高的预测精度,适用于装备保障资源需求预测。
Because of the difficulty of predicting which resources and services will be needed to meet the supporting resources requirements in modern war,this paper firstly analyzes the factors which effect supporting resources requirements systematically and selectes eight core factors according the fact,such as Mean Time Between Maintenance（MTBM） and Mean Time Between Repair（MTBR）.And then,apply GA-neural network in supporting resources requirement prediction,a model base on GA-BP neural network is proposed to predict the requirements.By using the data of requirements between 1980 and 2010 to train and predict,the predict model show fast convergence,high adaptability and preferable precision,and is suitable for supporting resources predict.
出处
《火力与指挥控制》
CSCD
北大核心
2013年第8期72-75,共4页
Fire Control & Command Control
关键词
保障资源
遗传算法
神经网络
需求预测
网络优化
supporting resources
genetic algorithm
neural network
requirements prediction
network optimize
