【作者】 王娜；

【导师】 赵玲；

【作者基本信息】 北京邮电大学 ， 管理科学与工程， 2010， 博士

【摘要】 虚拟产业集群能够突破地理限制实现网络范围的扩大和网络上的虚拟集聚,在更大的范围内整合各种资源,与传统产业集群相比具有更强的开放性和灵活性,是传统产业集群提升竞争力的有效组织形式。复杂性科学的兴起为产业集群尤其是基于网络集聚的虚拟产业集群的研究提供了新的分析工具,这种工具能够克服传统分析方法的不足。本文以虚拟产业集群为研究对象,运用复杂网络的建模思想和分析方法,采用发现——建模——分析——仿真——控制——综合的研究思路,综合运用演化博弈、仿真模拟、统计学的研究方法,按照虚拟产业集群网络演化→虚拟产业集群内的各种动力学过程→虚拟产业集群风险控制的逻辑思路,系统的挖掘虚拟产业集群网络演化机制和主要结构特征,构建面向虚拟产业集群的网络演化模型,在此基础上研究虚拟产业集群的结构与集群功能的内在联系、以及群内主体和要素的各种动力学行为和过程,并进一步考察虚拟产业集群的风险问题,构建虚拟集群的风险参数体系和风险评价模型,为虚拟产业集群的风险诊断、预测与控制提供理论依据。1.从宏观和微观两个方面分析了虚拟产业集群网络演化的动力因素,并提出了虚拟产业集群网络演化的关键机制,包括：增长、局部择优连接、动态演化、资源约束、优胜劣汰等,在此基础上运用复杂网络方法构建了虚拟产业集群网络演化模型,对其拓扑特征进行了测度,分析了该网络的度分布、集聚系数和平均路径长度等统计特征。2.利用演化博弈理论进行分析了不同地域和不同集群内的企业突破地理限制加入虚拟产业集群的动机和决策行为,综合考虑了影响企业入群的多种因素,如：企业入群的收益增量、企业入群的成本增量、企业的入群偏好、企业对未来的预期等,设计了企业入群的演化博弈模型。使用微分方程稳定性理论求出了该博弈模型的稳定均衡解,并画出演化博弈的相图进行分析,在分析的基础上提出了企业入群的激励机制。3.考虑到技术标准的形成过程对于产业市场整合和集群虚拟化发展的重要作用,本文采用复杂网络上的阈值模型研究了虚拟产业集群技术标准形成和扩散的过程。在建模中,重点考察了集群的网络结构、企业主体的异质性、网络效应等对标准扩散的影响。通过建模和仿真分析,得出如下结论：竞争性的技术标准在市场中的分布状况既与企业所处的网络结构有关,也和企业之间差异程度相关。同时,直接网络效应和间接网络效应在标准形成的过程中的作用完全不同,其中直接网络效应对某种标准形成赢家通吃的结果起阻碍作用,而间接网络效应却起促进作用。此外,初始安装基础对市场均衡状态的影响是不确定的。4.虚拟产业集群中最重要的动力学过程之一是知识的流动和扩散过程,本文通过构建知识网络模型研究了虚拟集群中知识的流动以及知识创新的过程,形象的分析了集群中知识网络的性质、知识数量和种类的变化过程以及创新对企业生产过程的影响。5.分析了虚拟产业集群可能发生的多种风险,并提出了一种新的风险类型——相继故障风险。利用复杂网络上的负荷——容量模型来研究虚拟产业集群的相继故障风险问题。重点讨论了虚拟产业集群网络的结构特征、企业自身特征分布的差异性和集群系统风险之间的关系。根据分析的结果,提出了相关政策建议。构建了虚拟产业集群的风险参数指标体系,并建立了集群风险评价模型。本文研究了虚拟产业集群的网络演化、资源流动、标准形成、创新扩散等过程,所采用的研究思路构建了产业集群的复杂网络分析框架,所得出的结论对于促进我国产业结构升级、整合或开发全球市场、增强我国产业整体竞争能力、增加产业集群的持续创新能力具有重要意义。更多还原

【Abstract】 Virtual Industry Cluster with special characteristics is an effective organizational form to enhance the competitiveness of traditional industries, especially high-tech and creative industry. The rise of complexity science provides a new analysis tool to study of industry clusters and can overcome the deficiencies of traditional methods. In the background of global economic integration, it is important for the management innovation on industry clusters to study the evolustion of virtual industry clusters. The main contents and innovations can be summarized in the following five areas:1. The evolutional factors of virtual cluster network are proposed from the macro and micro aspects and the key mechanisms of virtual cluster network evolution are presented. Then a network evolution model about virtual cluster network is build, and its topological features included the degree distribution, clustering coefficient and average path length and other statistical characteristics are discussed.2. We analyze the motivation and the decision-making behaviors of enterprises overcoming geographic limitations to join the virtual cluster by evolutionary game theory. A model of evolutionary game is designed and the EES solutions of the model are obtained by using the theory of differential equations. Based on the analysis we put forward the incentives mechanism for enterprises joining a cluster.3. Taking into account the technical standards-making process playing important roles for industry integration and the development of virtual clusters, our paper discussed process of the formation and diffusion of technical standards in virtual industry clusters by the threshold model in complex networks. We taking into account the affects of network structure of the cluster, heterogeneity of enterprise, network effects on the standard diffusion in the model. The outcomes of simulations indicate that standards diffusion is related with the interaction among actors or enterprises network structure. The direct network effects and indirect network effects play an opposite role in standards diffusion. A direct network effect is a brake and an indirect network effect is the impetus in diffusion process. The advantage of initial market share does not always mean final success and getting winner-take-all.4. As one of the most important dynamics processes is knowledge flows in virtual industry cluster, we explore knowledge flows and knowledge creation process by constructing knowledge flows network in the virtual cluster. The nature of knowledge networks, the change process in the number and types of knowledge, and the relationship between innovation and production are discussed.5. There are structural, cyclical and network risk in virtual clusters as other types of industry clusters. When the risk occurs, it often expresses a process of cascading failures. This article studies the evolutional process of cascading failures in virtual clusters based on Load-Capacity Model of complex network. Especially, the relations of network structural characteristics of virtual clusters, distribution characteristics of virtual enterprises and system risks are explored. In the last, the risk prevention mechanism is proposed. According to the analysis of the results, relevant policies are recommended. In order to detect risk we construct a parameters system about the risks of virtual industry cluster and further establish an assessment model of risks.The research thinking and research process build an analysis framework of industry cluster in complex network. The conclusions have important significances in promoting China’s industry structure upgrade, enhancing the overall competitiveness of our industry and increasing the sustainable innovative ability of industry clusters.更多还原