Equilibrium Customer Strategies in Markovian Queues with Setup Times and Partial Failures

This paper considers the equilibrium behavior of customers in a Markovian queue with setup times and partial failures, where the reactivated server must go through a period of setup time to reach the normal working state and a failure can occur at any time during the normal service. When a partial failure occurs, the server continues to serve the customers on spot at a low rate and does not admit a new arrival. Once the system becomes empty, an exponential repair time starts. Assuming that all the customers have the option of joining or balking based on a linear reward-cost structure, the authors analyze the equilibrium strategies of the customers and the average social benefits of the system in the fully observable case and the partially observable case, respectively. And on this basis, the effect of several parameters on customers' strategic behavior is presented by some numerical examples.     

基于MATLAB矩阵运算的大整数乘法设计与实现

大整数运算在信息安全、数学验证、基因工程等领域有着广泛的应用,设计有效的方案提高运算效率成为学者关注的热点。大整数乘法是大整数运算中的核心运算,对如何提高大整数乘法运算效率进行了分析总结,并利用MATLAB矩阵运算结合格子乘法等算法进行了设计与实现。实验表明通过MATLAB矩阵运算进行大整数乘法运算能有效的提高运算效率。     

神经网络与图论

本文讨论了神经网络与图论之间的相互关系：神经网络的深入研究与发展离不开图论，神经网络的深入发展必将对图论中许多问题的进一步研究产生重要的影响。文中论述了图论在神经网络研究中的作用、地位及应用现状，同时概述了神经网络在图论中的应用情况。     

Some synchronization criteria for general complex networks with coupling delays

A general complex delayed dynamical network model with asymmetric coupling matrix is considered in this paper. For reducing the conservativeness of synchronization criteria, several novel synchronization stability conditions are presented by using delay decomposition methods. Numerical examples which are widely used to study delay-dependent synchronization stability are given to illustrate the effectiveness of the proposed methods.     

<Emphasis Type="Italic">H</Emphasis><Subscript>∞</Subscript> control for stochastic systems with Poisson jumps

This paper discusses the H _∞ control problem for a class of linear stochastic systems driven by both Brownian motion and Poisson jumps. The authors give the basic theory about stabilities for such systems, including internal stability and external stability, which enables to prove the bounded real lemma for the systems. By means of Riccati equations, infinite horizon linear stochastic state-feedback H _∞ control design is also extended to such systems.     

基于突变理论的员工主动离职行为定性模拟——以中国经济转型期的“离职潮”为例

刘易斯拐点的到来导致中国劳动力成本大幅上升,这使得如何避免员工主动离职成为保证企业人力资源稳定的核心问题。针对离职行为这一复杂系统,从实证研究与突变理论相结合的全新视角,构建了员工主动离职的突变概念模型。在此基础上引入定性模拟算法,提出了突变函数在离职行为研究中的参数估计方法,并通过长三角地区某企业的实际案例验证了该方法的合理性及鲁棒性。最后,根据经济转型期的离职潮,对员工主动离职的定性-定量混合突变模型进行虚拟实验分析。实验结果从精确测定员工离职的预警区、临界区和突变区及复杂经济环境与员工离职的关系,以及企业微观层面和政府宏观层面的控制策略等方面,为员工主动离职问题提供了决策支持。     

基于矢量约束的随机特征点选取算法

无人机利用视觉在未知区域自主着降时,提取的特征点具有数量多、随机性强等特点。针对利用随机特征点进行位姿估计精度低、稳定性差的问题,提出一种基于矢量约束的随机特征点选取算法。首先通过分析位姿估计方程可知,特征点地理坐标是影响方程组求解精度的重要因素;然后在引入矢量角均分度、矢量模值均值及矢量模值最大值三项约束指标基础上,制定了一种基于矢量约束的特征点选取策略;最后利用正交迭代算法对所选取的特征点进行位姿估计精度验证。实验结果表明,提出的算法相比于随机选取的特征点进行相对位姿估计精度更高,鲁棒性更强。     

Optimal combinational of quota-share and stop-loss reinsurance contracts under VaR and CTE with a constrained reinsurance premium

This paper considers the problem of minimizing the VaR and CTE of an insurer’s retained risk by controlling the combinational quota-share and stop-loss reinsurance strategy. With a constrained reinsurance premium, the authors give the explicit reinsurance forms and the minimal VaR and CTE of retained risk in the case of quota-share after stop-loss reinsurance and the case of stop-loss after-quota-share reinsurance respectively. Finally, the authors conclude that the quota-share after stop-loss is a better reinsurance strategy than stop-loss after quota-share to minimize the VaR and CTE with a same constrained reinsurance premium. And the pure stop-loss reinsurance is preferred for an insurer with a high level regulatory requirement.