双旋转货架拣货作业优化设计

本文讨论了由一个拣货员负责的双旋转货架系统,目标是寻找一种最优拣货顺序使得总拣货时间。该问题被证明为强NP-难问题,文章提出了启发式算法寻找最优拣货顺序,并对相关模型及算法进行模拟测试以证明其有效性,为方法选择与应用提供了依据。     

多元随机风险传染模型及沪铜场内外风险传染实证

为了更合适的度量金融市场间风险传染整体效果,本文通过理论分析构造了风险传染方程,其风险传染项相比已有的系数风险传染项与协方差时变风险传染项更具实际意义,并借鉴多元GARCH建模原理建立了结合均值溢出、波动溢出与风险传染项的多元随机风险传染模型,设计了模型的MCMC迭代求解算法,满足解的完整性。最后,运用模型对沪铜场内外风险传染现象进行了实证,实证结果不仅验证了一系列已有研究结论,同时还给出了一些符合期货实情的新结论,如金融市场间风险传染类似金融市场波动存在集聚效应、沪铜与沪铝市场存在风险传染交替变化现象、市场行情的变化能提前反映于风险传染效果中等。这也充分表明新模型的有效性、实用性以及优越性。     

基于多分类器动态选择与成本敏感优化集成的电信客户流失预测研究

针对不同样本在特征空间中具有不同的区域特性和不同分类算法之间的预测互补性,在电信客户流失预测理论基础上,融合多分类器动态集成理论和成本敏感学习理论,建立了电信客户流失多分类器集成预测的利润函数,并提出了一类新的基于多分类器动态选择与成本敏感优化集成的电信客户流失预测模型.首先使用K均值聚类法聚类训练样本成多个分区;接着使用NaiveBayes算法、多层感知机算法和J48算法在各分区样本上构建客户流失预测子分类器;最后使用改进人工鱼群算法分别对各分区的子分类器进行成本敏感优化集成.实验结果表明,所提出的基于多分类器动态选择与成本敏感优化集成模型的分类性能不仅优于由训练集全体样本所构建的3个单模型,也优于基于改进人工鱼群算法优化集成这3个单模型而得到的集成模型.     

Using a Grey model optimized by Differential Evolution algorithm to forecast the per capita annual net income of rural households in China

China is a major developing country where farmers account for over 57% of the population. Thus, promoting a rural economy is crucial if the Chinese government is to improve the quality of life of the nation as a whole. To frame scientific and effective rural policy or economic plans, it is useful and necessary for the government to predict the income of rural households. However, making such a prediction is challenging because rural households income is influenced by many factors, such as natural disasters. Based on the Grey Theory and the Differential Evolution (DE) algorithm, this study first developed a high-precision hybrid model, DE-GM(1,1) to forecast the per capita annual net income of rural households in China. By applying the DE algorithm to the optimization of the parameter λ, which was generally set equal to 0.5 in GM(1,1), we obtained more accurate forecasting results. Furthermore, the DE-Rolling-GM(1,1) was constructed by introducing the Rolling Mechanism. By analyzing the historical data of per capita annual net income of rural households in China from 1991 to 2008, we found that DE-Rolling-GM(1,1) can significantly improve the prediction precision when compared to traditional models.     

谈Dijkstra的银行家算法

阐述了死锁的形成原因与必要条件,介绍了银行家算法的功能与数据结构,分析了银行家算法法的思想与实质。并用VB语言对银行家算法进行了模拟实现。     

基于多目标二层规划的服务供应链服务能力协同决策模型

为了解决服务供应链系统运作过程中存在的服务能力建设过量或不足问题,本文构建了以服务集成商作为服务供应链整体计划、控制和协调中心的上层计划者、各服务供应商作为具有相对自主权的下层计划者的服务供应链协同决策机制,应用多目标二层规划方法建立了服务供应链服务能力分配的优化模型,同时给出了模型的具体算法框图和流程。最后,通过实证算例对服务能力协同决策优化模型进行了计算和有效性验证。     

A comparative study of legitimation strategies in hybrid regimes

Despite the growing body of research on ‘hybrid regimes’, few studies address the issue of their domestic legitimacy. Targeting this gap in the literature, this article explores the legitimation strategies of three hybrid regimes around the globe: Russia, Venezuela, and Seychelles. Although these countries differ markedly in almost every aspect that can be thought of, the political systems of all three cases combine formally democratic institutions with authoritarian political dynamics. The qualitative, comparative analysis presented in this article uncovers a number of remarkable similarities between the regimes’ respective legitimization strategies. However, while the strategies for engendering legitimacy are similar across the cases, the content of these strategies is different: the Russian leadership mainly relies on preserving order and nationalism, the Venezuelan regime employs a more populist strategy, and the Seychellois regime uses a more personal and particularistic approach. Our findings not only provide insights into the mechanisms hybrid regimes use to consolidate their authority, but also highlight important differences and similarities between hybrid regimes around the world.     

震后应急物资配送的模糊定位-路径问题研究

从系统集成优化的角度研究震后应急物资配送的一种新模糊定位-路径问题(LRP),综合考虑救灾点所在地理位置和地形导致的应急车辆行驶时间的随机性、救灾点应急物资需求量的不确定性与应急物资配送的时间紧迫性,以应急物资总运达时间最短与总配送成本最小为目标,构建一个基于机会约束规划的多目标模糊LRP优化模型,并根据模型的特点设计了一种混合免疫遗传算法予以求解。最后,通过算例验证了本文方法能有效解决震后应急物资配送的模糊多目标LRP,实现了震后应急物流中心定位和应急车辆路径规划的联合决策。     

基于核和灰度的灰色异构数据代数运算法则及其应用

为解决灰色异构数据的建模问题,应用"核和灰度"对灰色异构数据代数运算法则及其性质展开研究。将灰信息表征为"核和灰度",通过"核"将灰色异构数据代数运算转换为实数之间代数运算,根据灰度不减公理确定运算结果之灰度,在此基础上构建灰色异构数据的代数运算法则,并将该法则应用于灰色异构数据预测模型的构建及空气质量指数(AQI)的预测。研究成果对丰富与完善灰色系统基础理论具有积极意义。     

基于混合蚁群算法的生产系统设施规划问题研究

设施规划问题主要研究生产设备的布局规划,从而减小厂区内的物料搬运成本。一个有效的设施规划有利于生产过程中整体运作效率的提高。随着市场竞争的日趋激烈,市场环境处于不断的变化之中,制造企业需不断对设施布局进行重新规划来适应不断变化的市场环境对产品需求量的影响,并达到降低成本的目的。这一问题便需要用多阶段设施规划（MFLP）的方法来解决。本文提出了一种改进的混和蚁群算法（HACO）来解决带有财务预算约束的多阶段设施规划问题,并将此方法与其他一些典型的启发式算法进行了对比分析。结果表明,本文提出的HACO算法是求解带有财务预算约束的MFLP问题的一种有效的方法。