BP神经网络的飞机机体研制费用估算

提出一种应用人工神经网络预测飞机机体研制费用的方法.该方法利用偏最小二乘法对飞机性能参数进行了主成分提取,用这些主成分作为BP网络的输入变量,建立了一种基于BP神经网络的飞机研制费用的估算模型,应用该模型对典型的飞机机体研制费用进行了预测.预测结果表明,该方法是有效的、可行的.     

武器系统模糊寿命周期费用模型

武器系统的寿命周期费用建模较多采用参数法,而参数法中最常用的是最小二乘回归.考虑费用统计数据的模糊性,提出用模糊最小二乘回归来建立武器系统模糊寿命周期费用模型,并结合实例对武器系统寿命周期费用进行了分析.结果表明,这种方法能达到令人满意的拟合精度,具有实用价值.     

THE EFFECTS OF THE WORLD DEFENSE INDUSTRY AND US MILITARY AID TO ISRAEL ON THE ISRAELI DEFENSE INDUSTRY: A DIFFERENTIATED PRODUCTS MODEL

This paper models the interactions between the defense industry market structure and the defense needs of Israel, the USA and Western Europe, which produce several heterogeneous defense goods. The model specifies that the defense industries of the US and Europe are ‘large’ while that of Israel is ‘small’. The US military aid to Israel is also an integral part of the model. The results show that net defense costs of Israel are minimal when the number of its defense firms is one. The model predicts that an increase in US military aid reduces Israel’s government expenditure, its defense industry’s profits and its net defense costs.     

The structure of the defense industry and the security needs of the country: a differentiated products model

This paper models the interactions between the defense needs of the USA and Western Europe, which produce several heterogeneous defense goods, and the defense industry market structure. The results show that net defense costs of the USA and Europe are lower when the number of defense firms in each arms‐producing country is small and when the world prices of the defense goods are high. The model predicts that the increase in world prices will crowd‐out countries in the developing world from the market for modern weapon systems and may force them to develop and use ‘cheap and dirty’ weapon systems.     

Estimating defence budget saving from disarmament: the United States’ case

Probing the technology in the production of US national defence by using a dynamic cost‐function model with adjustment costs, this paper evaluates the effect of reducing the level of national defence on the defence budget saving. Our inquiry involves estimating the defence production structure without output data for non‐market goods that are normally unavailable. Our findings include: (i) the United States behaves rationally to minimize cost in the production of national defence; (ii) the adjustment costs are larger in disarmament than in military build‐up; (iii) due to the adjustment costs peculiar to disarmament, the defence budget saving from disarmament appears small, but cutbacks allow great savings on the defence budget.     

一种面向宽带复杂信号侦察的可配置计算应用框架

可配置计算系统的优势与系统设计过程的复杂性是一对矛盾,可配置计算框架为此提供了有益的思路。但受限于适用范围与通用性考虑,已有的特别是商业框架仅支持系统的静态重构。本文针对宽带信号侦察这一特定应用领域,提出了一种功能级可配置计算应用框架BSRRCS-RCAF,该框架将领域应用中具有共性的计算与功能模块抽取和集中起来,形成二进制比特流配置文件库,并提供构建此类系统的公共服务体系。基于公共服务体系以及配置文件库,当需要修改或者重新构建宽带信号侦察系统时,只需在现有框架下动态调用配置组件库,或者遵循框架的统一设计接口完善补充组件库中的特定功能部分,整个系统设计框架将得到方便复用,便于实现基于快速重构的多种系统功能。     

基于栅格分层的逐栅格汇流算法并行化研究

分布式水文模型中的逐栅格汇流算法计算量大,需要借助并行计算以满足大流域长历时模拟的要求。针对目前鲜有对基于隐式有限差分的逐栅格汇流算法进行并行计算研究的情况,基于栅格分层的思想提出一种适用于共享内存并行计算环境的逐栅格汇流并行算法。该算法首先根据流向进行栅格分层,使同一层中栅格的计算相互独立,然后将同一层中栅格的计算任务分配到多个计算单元并行计算。采用C++编程语言与OpenMP并行编程库实现了该算法,并选择河北省清水河流域为实验区,在不同数据规模(30m、90m、270m分辨率)、不同核数(2～20个)以及不同栅格分层方法的情况下对算法性能进行了测试。实验结果表明本文提出的并行算法具有较好的加速比和并行效率,且并行效率随数据规模的增大而增大。栅格分层方法对算法并行性能有明显影响,从上到下的分层方法比从下到上的方法具有更高的并行效率。     

面向混部云失败批处理作业的预测算法

为了降低混部云失败批处理作业的风险,使用K-means聚类算法将批处理作业分为四类,在分类的基础上提出了二层嵌套分类模型(two-layer nested classification model, TLNM),实现了基于TLNM的预测算法。基于Ali Trace 2018数据集上的实验结果表明,该算法的接受者操作特性(receiver operating characteristic, ROC)曲线明显优于其他常用分类器,ROC曲线下面积(即AUC)可以达到0.978,表明该算法具有良好的分类性能。同时召回率可以达到0.951,通过混淆矩阵可以看出TLNM算法能够准确预测出执行失败的批处理作业。     

软计算在污水处理过程控制中的应用

阐明了软计算技术在污水处理过程中应用的必要性、可行性及对行业所带来的社会效益和理论意义,着重分析了其在污水处理过程中的应用情况,简要探讨了软计算技术今后应深入研究的问题和方向,指出将软计算算法和智能控制策略结合起来有利于提高模型预测精度,提高过程控制系统动态响应能力。     

ET:一种能耗有效的高性能嵌入式处理器

随着标准和算法的不断演进,高端嵌入式应用对性能和能耗提出了越来越高的要求.然而,能耗问题成为将VLSI潜力转换为实际应用需求的最大挑战,基于此,提出ET( Embedded Tera-scale Computing)处理器设计.ET以众多轻量级处理器(称为小核)来搭建目标处理器,每个小核都是一个基于显式数据和指令管理的...