空陆攻防博弈的动态武器目标分配

大规模作战具有高动态、非完全信息和不确定性,在分析归纳目前解决动态武器目标分配问题的一系列方法的基础上,尝试构建基于双方动态博弈的攻防对抗综合数学模型,并利用纳什均衡和帕累托最优算法进行分阶段求解。结果表明,该数学模型和博弈论方法结合能够有效解决武器目标动态分配问题。     

Purchase and retrieval competition for seasonal produce

We study the competition problem of purchase and multiretrieval of perishable seasonal produce, where wholesalers purchase and stock their products in the first period, and then retrieve and sell them in subsequent periods. We first consider the duopoly case and assume that the prices are exogenous and fluctuate. In each period, after the price realization, the wholesalers retrieve some stock from their warehouses to satisfy their demands. One wholesaler's unsatisfied customers can switch to another and be satisfied by its left retrieved products. Any unsold retrieved stock has no salvage value and any unsatisfied demand is lost. The unretrieved stock is carried to the next period at a perishable rate. The wholesalers compete for the substitute demand by determining their own purchase and retrieval quantities. We show the existence and uniqueness of a pure-strategy Nash equilibrium, and that the Nash equilibrium strategy has the simple “sell-down-to” structure. We also consider the general N-person game and show the existence of the Nash equilibrium, and characterize the structure of the equilibrium strategy for the symmetric case. In addition, we consider the case with endogenous prices, and show that the problem reduces to a repeated newsvendor game with price and inventory competition. We derive the conditions under which a unique Nash equilibrium exists and characterize the equilibrium strategy. Finally, we conduct numerical studies to examine the impacts of the model parameters on the equilibrium outcomes and to generate managerial insights.     

状态空间表达下控制系统的稳态误差

经典控制理论对系统稳态误差的讨论从传递函数入手,重点是对系统开环传递函数的研究,当系统的输入是任意函数时,引入动态误差系数方法来研究稳态误差,但是当输入具有高阶导数时,动态误差系数将很难得到.现代控制理论中的状态空间表达下求系统的稳态误差很好地解决了这个问题.利用矩阵之间的运算来表示动态误差系数,并且可以得到任意输入下的稳态误差值,在线性定常系统下的推导结果还可以适用于线性时变系统,具有一定的普遍性.     

军事欺骗战法分析的数学方法-应用*

针对军事欺骗战法定量分析困难的问题,采用基于贝叶斯统计推断和博弈均衡分析的数学方法,在理论上为解决此难题提供了一种可行的方法.该方法首先定义基本概率矩阵,然后根据贝叶斯原理为交战双方构造用于统计推断的主观概率矩阵,并在由两个主观概率矩阵元素构成的偏序链与实施军事欺骗机会之间建立对应关系,再将此关系推广到基本概率矩阵,即是否存在实施军事欺骗的机会将由基本概率矩阵元素构成的偏序链决定.对理论成果的初步应用表明:通过分析由战场情报信息构成的基本概率矩阵的偏序链,可以科学地运用军事欺骗和反军事欺骗战法.     

Group support for political violence: The role of emotions and expressive choice in creating conflict or providing peace

This paper provides a rationale for group support for political violence which does not provide a material benefit. Rabin’s (1993) theory of fairness is adopted to demonstrate that although group violence may not be a Nash equilibrium it may be a fairness equilibrium in a game containing psychological payoffs. For this to happen the material stakes must be perceived as low and psychological payoffs are expressive. Although the material stakes are actually high, members of each group may choose expressively to support the use of violence because the probability of being decisive is low. The paper also considers the possibility of peace emerging as a fairness equilibrium. This can only happen if each group perceives the other as making some sacrifice in choosing peace.     

一类具有标准发生率的SI型传染病模型的全局稳定性

研究如下一类具有标准发生率的SI型传染病模型{ds/dt=rS（1-（S＋I）/K-β SI/（S＋I）, dI/dt=βSI/（S＋I）-dI。应用微分方程定性理论,给出了该系统地方病平衡点、无病平衡点和总人口消亡平衡点的全局渐近稳定的充分条件。     

李雅普诺夫运动稳定性与平衡状态稳定性的关系

稳定性是系统分析的重要内容 ,针对在系统稳定性分析中 ,运动稳定性和平衡状态稳定性之间常易于混淆的问题 ,详细分析了两种稳定性之间内在本质联系 ,给出了从一种稳定性过渡到另一种稳定性的方法     

模拟体系对抗价值链的网络同步模型

描述战斗毁伤"规模效应"的Lanchester模型默认作战体系内部是"完美的"。本文扬弃此假设,首次从作战体系内部"共识"涌现的视角提出了一个模拟体系对抗价值链的网络"同步效应"模型,研究了战场变化、个体/群体认知、自同步与对抗均衡等概念的网络拓扑和权值演化形式,并发现在对抗均衡条件下无标度网络具有"同步脆弱性"。     

大区域防空的进化群决策机制研究

大区域防空所涉及的区域广阔,因素众多,其作战计划的制定比较复杂。为解决此问题,提出了一种基于改进德尔菲法的具有进化学习功能的群决策机制,得到了纳什均衡解,使作战方案的制定更加合理,有助于防空作战效能的提高。     

非对称Colonel Blotto博弈模型下的多信道功率分配抗干扰

针对智能干扰条件下传输速率固定的通信系统多信道功率分配问题,建立了非对称Colonel Blotto博弈模型。在完全信息条件下,推导出了各种功率预算约束下通信方和干扰方的等效单信道最优功率分配策略,进而证明了通信方和干扰方存在唯一混合纳什均衡策略,并求得了纳什均衡收益。基于等效单信道最优功率分布,提出了一种多重扫描直接列元素交换算法,可以快速构建多信道混合功率分配矩阵,且相比于线性规划方法,可适应更多的信道数和更广的功率分布范围。通过数值仿真,验证了所提多信道混合功率分配矩阵构造算法的有效性及多信道功率分配策略的最优性。