Continuum versus discrete networks,graph Laplacians,and reproducing kernel Hilbert spaces

Motivated by applications to machine learning, we construct a reversible and irreducible Markov chain whose state space is a certain collection of measurable sets of a chosen l.c.h. space $\mathfrak{X}$. We study the resulting network (connected undirected graph), including transience, Royden and Riesz decompositions, and kernel factorization. We describe a construction for Hilbert spaces of signed measures which comes equipped with a new notion of reproducing kernels and there is a unique solution to a regularized optimization problem involving the approximation of $L^2$ functions by functions of finite energy. The latter has applications to machine learning (for Markov random fields, for example).     

Neural network modeling of hydrological systems: A review of implementation techniques

This paper presents a review of procedural steps and implementation techniques used in the development of artificial intelligence models, generally referred to as artificial neural networks (ANNs), within the water resources domain. It focusses on identifying different areas wherein ANNs have found application thereby elucidating its advantages and disadvantages as well as various challenges encountered in its use. Results from this review provide useful insights into how the performance of ANNs can be improved and potential areas of application that are yet to be explored in hydrological modeling. Recommendations for Resource Managers

• Development of integrated and hybrid artificial intelligent tools is critical to achieving improved forecasts in hydrological modeling studies.

• Further research into comprehending the internal mechanisms of neural networks is required to obtain a practical meaning of each network component deployed to solve real‐world problems.

• More robust optimization techniques and tools like differential evolution, particle swarm optimization and deep neural nets, are yet to be fully explored in the water resources analysis, and should be given more attention to enhance neural networks aptitude for modeling complex and nonlinear hydrological processes.

     

基于实际路网情境的配送车辆调度优化

在实际路网情境下结合车道数、车道宽度、路口信号灯设置等路网物理特性,构建了考虑综合交通阻抗的多车型车辆调度模型,提出了两阶段求解策略:第1阶段设计了改进A-star精确解算法用于计算客户时间距离矩阵;第2阶段针对实际路网的特征设计了混合模拟退火算法求解调度方案。以大连市某配送中心运营实例进行路网情境仿真试验,结果表明:改进A-star算法较改进Dijkstra算法具有更短的路径搜索时间;混合模拟退火算法求解结果较实际调度方案优化了13.1% 的综合成本;路网增流、区域拥堵和路段禁行三类路网情境均能对配送方案的车辆配置、路径选择、客户服务次序、作业时间和违约费用等5方面内容产生干扰,调度计划的制定需要详细考虑这些因素的变化。     

The (S−1,S) inventory model and its counterparts in queueing theory

We explore the relationship between the ($S?1,S$) inventory model and three well-known queueing models: the Erlang loss system, the machine-repair model and a two-node Jackson network. Exploiting this relationship allows us to obtain key performance measures of the ($S?1,S$) model, like the so-called virtual outdating time, the number of items on the shelf in steady state, the long-run rate of unsatisfied demands and the distribution of the empty shelf period.     

Optimal intervention under stress scenarios: A case of the Korean financial system

We consider optimal intervention methods under budget constraints when financial systems face economic shocks. We propose two policies formulated by mixed-integer linear programs where regulators inject cash into institutions. One is to minimize systemic losses, and the other is to minimize the number of defaulting institutions. Using publicly available data on the Korean financial system, we construct its entire network and apply stress scenarios to the system to compare the performances of intervention strategies and derive insights on their workings.     

面向远海岛礁群的双向物流网络规划

针对若干远海岛礁群系统封闭性的分布特点,将中心岛礁选址、各岛礁建设泊位数量和规模、仓库储存量、运输船型和数量以及航线设置、班期组织等作为规划内容,以整个运输网络系统总成本最低为原则,综合考虑了仓库存储理论,建立了兼顾“选址-库存-路径”的规划模型。结合所建模型特点,提出了双层并行搜索遗传算法,构建了以海上运输为基本交通方式的双向物流系统。最后,以我国南海为例进行建模并求解,采用不同算法并通过多组不同规模的算例进行演算比较,验证了所建模型和算法的有效性与合理性。     

基于ANP法的股权激励模式选择研究

本文对企业如何科学合理地选择股权激励模式问题进行研究探讨。不仅考虑到以往研究中的企业特征因素,同时也考虑了激励对象特征因素和外部实施环境因素,建立股权激励模式选择指标体系,并分析各指标间的依存关系,基于此构建了ANP网络结构模型,并选取两个代表性样本企业进行算例验证和应用,最终使得不同企业可根据自身的实际情况选择合适的股权激励模式,验证了该模型的合理性与可行性。     

基于复杂网络的能源互联网信息物理融合系统跨空间风险传递研究

能源互联网呈现物理信息深度融合的趋势,为电力系统管理研究定义了新的研究框架。作为一个先进的复杂系统,能源互联网信息物理系统(Energy Internet Cyber-Physical System,ECPS)在其发展过程中面临着一些新的挑战,其中一个就是耦合结构下的风险管理。本文结合复杂网络理论和风险传递理论,着重在拓扑层面分析了ECPS跨空间交互机理,并在此基础上定义了交互路径和交互系数;接着建立了ECPS跨空间风险传递模型,量化描述了风险的传递和演化过程,并进行了风险影响评估;最后,通过仿真实验分析了三种不同交互系数节点故障的风险传递过程和不同攻击模式下系统的崩溃过程。对仿真结果的讨论阐述了能源互联网风险跨空间传递的特点,为更深入地研究ECPS的风险管理提供了参考。     

基于荧光光谱技术的菜籽油氧化状态智能评价

菜籽油在加工及贮藏过程中，易受氧气、温度、光照等因素的影响，产生氧化酸败现象。为准确判断油脂氧化程度，实现不同氧化模式下菜籽油品质的快速判别，采用三维同步荧光光谱技术结合平行因子分析法及BP神经网络法建立菜籽油氧化状态的智能评价模型。以冷榨菜籽油为原料，将样品分别置于常温、Schaal烘箱、高温模式中氧化处理，期间采集菜籽油的三维同步荧光光谱数据及理化指标，当理化指标超出国标限定范围时，停止采集数据。结果表明，菜籽油中荧光物质在不同氧化模式中的演变规律呈显著差异，氧化温度对菜籽油荧光光谱有明显影响。常温氧化350 d与第1 d相比，菜籽油的特征荧光峰位置无变化，仅在激发波长Ex为620和660 nm附近荧光峰强度发生微弱变化；Schaal烘箱氧化26 d后，在激发波长Ex为620和660 nm附近荧光峰强度显著减弱，且在激发波长Ex为350～450 nm之间有新的荧光峰生成；高温氧化48 h后，Ex为620和660 nm处荧光峰消失，在Ex为400～550 nm处产生显著荧光峰，相对Schaal烘箱氧化，荧光波长发生一定程度红移，这是由于高温氧化过程中油脂氧化生成的物质稳定性较差引起的。利用平行因子分析法对三维同步荧光光谱数据进行分解获取有效的二维荧光光谱数据，当组分数为6，Δλ=60 nm时激发波长的载荷值最大，不同样品间差异最显著。选定Δλ=60 nm波段的二维荧光光谱数据用于智能评价，作为BP神经网络模型的输入值，以极性组分作为模型输出值，分别对菜籽油三种氧化模式数据建模训练。实验结果表明，三种氧化模式对应的训练集、验证集、测试集模型相关系数r均能达到0.9以上，其中常温氧化模式中验证集及测试集模型的相关系数r为1，输出值与目标值较接近，模型的预测效果较好；综合三种氧化模式数据建模，对应训练集、验证集、测试集模型的相关系数分别为0.999，0.913和0.988，均方误差均较小，说明该模型能准确判断菜籽油的不同氧化状态。因此，三维同步荧光光谱技术结合平行因子分析法、BP神经网络法建立快速检测模型能实现菜籽油不同氧化状态的判别，为菜籽油的氧化程度的评价提供新方法，同时为其他食用油的品质评价提供参考。