【作者】 黄芳；

【导师】 鞠颂东；

【作者基本信息】 北京交通大学 ， 管理科学与工程， 2010， 博士

【摘要】 信息技术和网络经济的高速发展,物流运作无国界、全球化发展趋势,物流作用的经济效应得到高度关注。全球化作用下,企业处于开放的,综合性、复杂性与网络化的运作环境之中。物流企业通过资源的整合,形成制造企业、销售企业、用户、以及包含物流联盟组织、虚拟组织及其他物流组织形态在内的无边界的开放的网络。网络中各节点信息和资源可以充分共享,协同作业,为客户提供一体化物流服务。物流系统各环节紧密关联、彼此依赖,任何一个环节出现问题,都可能波及其他环节,这种运作特点的改变也致使物流网络的风险呈现多样化、复杂性的特征。物流网络这种无边界的、开放的、动态的特性,体现了物流网络运作宏观性特征。从微观层面看物流网络的运作组织和机构中,其经营管理活动发生了根本性变化。因此,基于网络环境下的物流运作风险研究意义重大。由于物流网络相关研究较新,对物流网络运作风险至今还没有统一的界定与认识。论文结合风险的一般定义以及物流网络的内涵,认为物流网络运作风险是指物流网络在运作过程中,由于不确定性因素的影响,导致成员企业的实际收益与预期收益发生偏差,而蒙受损失的机会或可能性。这里强调风险的主体是物流网络的所有参与者。从目标控制的角度出发,物流网络运作风险是组织偏离预定目标的可能性。论文对物流网络运作风险的研究重点是基于物流网络一体化运作过程中系统性风险以及网络各成员企业间协同性风险的风险评估、预警、控制的研究。在物流网络运作风险评估部分,论文采用了交互式人工智能的方法进行风险评估的研究。论文首先界定了风险识别、分析与评估的基本概念、方法以及流程等内容。其次,论文研究了物流网络运作风险评估的流程、指标、风险事项测度等内容,在此基础上,构建了基于交互式风险评估模型。最后,论文运用物流网络风险评估的理论与方法,对安德物流花城分公司的物流网络运作进行了风险评估分析。交互式人工智能的方法一方面能够极大提高评估速度,另一方面可以避免由于评估结果过分依赖于某个专家的个人判断,降低系统性偏差。在物流网络运作风险预警部分,论文采用了事例式推理(CBR)的方法进行风险预警的研究。论文首先界定了物流网络运作风险预警的概念、功能、结构及关键技术,在此基础上,构建了基于事例推理的人工神经网络模型,并对预警模型的结构、学习机制、分析算法和信号输出等进行了研究。最后,论文运用物流网络风险预警的理论与方法,对安德物流P分公司的物流网络运作进行了风险预警分析。人工神经网络采用分布式存储结构,学习能力强,容错性好,具有很强的鲁棒性,可以有效完成风险预警。在物流网络运作风险控制部分,论文分析了物流网络运作风险的成因,将COSO风险全面监控框架引入物流网络运作风险控制,建立三维的风险监控框架和体系,并给出了物流网络运作风险控制策略。COSO风险管理框架能够从整体上关注物流网络运作风险管理,也可以从目标类别、构成要素或主体单元的角度去理解和控制风险,平衡物流网络中成员企业之间的利益关系,实现不同层次的战略目标,有效控制物流网络运作的风险。本文中共有图49幅,表9个,参考文献155篇。更多还原

【Abstract】 Because of the role of information technology, businesses are in open, comprehensive, complex and network-based operating environment. Through the integration of resources, logistics alliance organizations, virtual organizations and other logistics organizations are formed by combined manufacturing companies, marketing companies, and end-users. In this network, information and resources can be fully shared, and node enterprises developed from "parallel", and "do not intersect" into synergies. Each node enterprise are dynamically restructuring and forming of a number of different types of service organizations together to provide customers with integrated logistics services. Logistics enterprise’s operation and management fundamental changes in such an open, free and borderless network of organizations and institutions. Logistics network is closely related to each other, and any one of the problems may spread to other sectors. Therefore, the logistics network risk research based on the network nodes are great significance.As logistics network research relatively new, there is still no uniform definition and understanding about logistics network operation risks. According to the general definition of risk and logistics network, the dissertation define that logistics network of operational risks is the process of logistics networks in operation, due to uncertainties that may cause the actual members of the corporate earnings and the expected return deviations occur, but suffered losses opportunities or possibilities. The main risk is all the participants of logistics network. From the target control point of view, the logistics network operation risks is the possibility that the organization deviation from the intended target. The research of logistics network operation risks focus on risk assessment, early warning and control about logistics network integration operation risks and networking various members’synergistic risks.In the logistics network operation risk assessment section, the dissertation using an interactive artificial intelligence approach to risk assessment. First, the dissertation defines the basic concepts for risk identification, analysis and assessment, methods, and procedures and so on. Second, the dissertation studies the risk assessment process, indicators, risk measurement and other contents about the logistics network operation, on this basis, sets an interactive-based risk assessment model. Finally, the dissertation uses the theory and methods, about logistics network risk assessment, conduct a risk assessment analysis of logistics network operation on Andre logistics Huacheng branch. On the one hand, interactive artificial intelligence approach can greatly improve the assessment speed, The other hand, the assessment results excessive reliance on an expert’s personal judgments can be avoided and reduce systemic bias.In the section of logistics network operation risk early-warning, the dissertation uses case-based reasoning (CBR) methods to research early warning of risk. first,the dissertation defines early warning concept, functions, structure and key technologies of logistics network operation risks, On this basis, build artificial neural network model based on case-based reasoning, the structure of early-warning model, learning mechanisms, analysis of algorithms and signal output are studied As well. Finally, the dissertation using theories and methods of logistics network risk early warning, analyzes the early warning risks of logistics network operation on Andre logistics P branch. Artificial neural network using a distributed storage architecture, learning ability, good fault tolerance and strong robustness, it can complete the risk early warning effectively.In the section of risk control of logistics network operation, the dissertation analyzes the causes of logistics network operation risk,COSO risk management framework be introduced to risk control of logistics network operation, establishes a three-dimensional framework and system of risk monitoring, put forwards risk control strategies of logistics network operation.COSO risk management framework concerns the whole risk management of logistics network operation,Can understand and control risks from the target category, constituent elements and the main unit, Balances the interests relationship between the of member enterprises in logistics network, achieves the strategic objectives at different levels and control the risks of logistics network operation effectively.In addition, there are 49 figures and 9 tables in the dissertation. We also have referred to 155 reference documents.更多还原