Impulsive synchronization of time-varying dynamical network

Synchronization of time-varying dynamical network is investigated via impulsive control. Based on the Lyapunov function method and stability theory of impulsive differential equation, a synchronization criterion with respect to the system parameters and the impulsive gains and intervals is analytically derived. Further, an adaptive strategy is introduced for designing unified impulsive controllers, with a corresponding synchronization criterion derived. In this proposed adaptive control scheme, the impulsive instants adjust themselves to the needed values as time goes on, and an algorithm for determining the impulsive instants is provided and evaluated. The derived theoretical results are illustrated to be effective by several numerical examples.     

优化的径向基-循环子空间网络为药物定量构效关系建模

径向基．循环子空间回归(RBF-CSR)网络,保留了径向基-偏最小二乘(RBF—PLS)网络的优点,且可在更广的范围内选择最优模型,但仍存在着参数难以确定,计算量大等问题。对此,本研究兼顾网络模型的拟合与预测性能,采用具有高效全局搜优能力的优进遗传算法(EGA)优化网络参数,构建为EGA-RBF-CSR方法,并将其成功应用于苯乙酰胺类除草剂的构效关系(QSAR)建模,效果良好,显示出很强的学习能力,所建模型具有良好的预报性能和稳定性,并优于其他方法。     

零售商价格领导权结构下考虑专利保护的闭环供应链定价策略

很多零售商常处于其所在供应链的主导地位,针对零售商具有价格领导权并且考虑专利保护的闭环供应链进行分析,运用博弈理论得到了零售商、原制造商和再制造商的最优策略,以及最优策略关于再制造成本节约的变化规律。研究表明:随着再制造成本节约的增加,最优单位专利许可费和废旧产品回收价格逐步增加,最优产品批发价格和销售价格逐步降低,并通过数值算例表明:零售商、再制造商和供应链系统的利润逐步增加,原制造商的利润逐步减少。     

基于成交量分解模型的改进VWAP策略

传统VWAP(交易量加权平均价格)策略通过拆分大额委托订单,跟踪市场成交均价,达到最小化冲击成本的目的,而准确预测成交量日内分布是运用VWAP策略的关键。通过详细考察现有的改进VWAP策略中成交量预测模型的建模方式和预测结果,发现由于无法分离成交量日内周期结构,现有模型样本依赖性较大且难以适用于多数股票。因此,本文从个股与市场成交量变化趋势的关系角度出发,推导个股成交量与市场趋势的关系,通过构造个股成交量关于市场因素的因子载荷,将日内成交量分解为市场共同部分和个股特殊部分,预测成交量日内分布并构建动态VWAP策略。实证结果表明新的成交量分解模型可以有效分离个股的成交量日内周期结构,在此基础上构造的改进VWAP策略不仅具有较为广泛的适用性,且跟踪误差减少幅度比现阶段同类型的改进VWAP策略更大,能更好的降低市场冲击成本。     

基于种群自适应调整的多目标差分进化算法

为提高已有多目标进化算法在求解复杂多目标优化问题上的收敛性和解集分布性,提出一种基于种群自适应调整的多目标差分进化算法。该算法设计一个种群扩增策略,它在决策空间生成一些新个体帮助搜索更优的非支配解;设计了一个种群收缩策略,它依据对非支配解集的贡献程度淘汰较差的个体以减少计算负荷,并预留一些空间给新的带有种群多样性的扰动个体;引入精英学习策略,防止算法陷入局部收敛。通过典型的多目标优化函数对算法进行测试验证,结果表明所提算法相对于其他算法具有明显的优势,其性能优越,能够在保证良好收敛性的同时,使获得的Pareto最优解集具有更均匀的分布性和更广的覆盖范围,尤其适合于高维复杂多目标优化问题的求解。     

基于非对称演化博弈的京冀雾霾协同治理联盟稳定性分析

当前,雾霾已成为制约首都经济、社会发展的首要环境问题。本文基于生态补偿及生态索赔的视角,构建了京冀雾霾治理联盟的非对称演化博弈模型,分析了京冀雾霾治理联盟的稳定性,并得到最优策略组合。研究结果表明:在不实行行政干预的情况下,由于成本因素及当地政府生态索赔措施的疲软,企业不会自主选择节能减排策略。合理的生态补偿标准对于企业采取节能减排策略至关重要。当两地生态索赔的总值低于企业节能减排的成本时,无益于企业采取节能减排策略,且生态补偿的额度过高会使得企业宁愿索赔不治污。只有两地生态索赔的总值高于企业节能减排成本,企业才会采取节能减排策略,并且实施节能减排策略不会被生态补偿的额度撼动。     

灰色市场条件下基于收益共享契约的供应链协调

考虑灰色市场条件下由一个制造商和两个分别处于不同国家的分销商构成的供应链,分别分析了完全分散化、部分分散化和集中化决策下的供应链定价策略。考察了各种决策模式下,消费者对灰市产品价值的认可程度对制造商定价策略的影响。通过比较得出分散化和部分分散化决策下的最优定价策略偏离集中化决策下的最优定价策略。并以供应链集中化决策下的利润为基准,通过引入收益共享契约针对灰色市场条件下的完全分散化和部分分散化供应链进行协调,并给出保证供应链节点企业达到帕累托改进的收益分享系数取值范围。最后通过数值分析给出了消费者对灰市产品价值的认可程度与供应链利润、灰市产品销量及收益分享系数之间的关系。     

碳排放交易机制下的供应链运作策略及协调研究

碳排放权交易是控制和减少碳排放的有效工具,同时也使得企业的运作成本增加,供应链的决策管理更复杂。考虑由零售商和制造商组成的二阶段供应链,研究不同碳排放权交易政策(供应链成员内部碳排放权交易、外部市场碳排放权交易)下基于数量折扣契约的供应链协调问题,并与无碳排放约束的情形进行比较。研究结果表明:在不同碳排放权交易政策下,数量折扣契约能够实现供应链的协调;无碳排放约束的供应链最优订货量大于外部碳排放权交易的供应链最优订货量,成员内部碳排放交易下的供应链最优订货量和供应链期望利润随着碳排放配额的增加最终等于无碳排放约束下的供应链最优订货量和供应链期望利润;外部市场碳排放交易政策下供应链的期望利润随着碳排放配额的增加而增加。     

High-pressure phase equilibria of binary and ternary mixtures of carbon dioxide,triglycerides and free fatty acids: Measurement and modeling with the GC-EOS

Vapor–liquid equilibria were measured for binary systems including carbon dioxide and free fatty acids such as oleic and linoleic acid, triglycerides as triolein or vegetable oils as sunflower oil, and modelled by the Group Contribution Equation of State (GC-EoS). Binary group parameters used in the calculation and prediction of CO₂–triglycerides systems were obtained from the literature but in the case of CO₂-free fatty acids, binary group parameters were adjusted through a parameterization strategy. At the end, a unique set of parameters could be established to model systems of vegetable oils at high pressures in the presence of CO₂. Moreover, a strategy to calculate the two-phase region in a ternary diagram was also studied. The good correspondence between theoretical and experimental results suggests that the strategy and the set of binary group parameters proposed in this work can be reliable predictive tools for using GC-EoS to describe systems involving vegetable oils with a composition based mainly in free fatty acids and triglycerides, such as those which can be encountered, for example, in oil deacidification processes using supercritical CO₂.     

A cops and robber game in multidimensional grids

We theoretically analyze the ‘cops and robber’ game for the first time in a multidimensional grid. It is shown that in an n-dimensional grid, at least n cops are necessary if one wants to catch the robber for all possible initial configurations. We also present a set of cop strategies for which n cops are provably sufficient to catch the robber. Further, we revisit the game in a two-dimensional grid and provide an independent proof of the fact that the robber can be caught even by a single cop under certain conditions.