【作者】 禹献云；

【导师】 曾德明；

【作者基本信息】 湖南大学 ， 企业管理， 2013， 博士

【摘要】 随着产品更新速度加快、市场竞争的日益激烈、技术创新对专业能力的需求愈加广泛使得产业内组织间的合作关系由松散的两点结构转变发展成为联系更为紧密的网络结构,协作研发网络由此应运而生。组织间的合作与科技创新资源共享必然推动知识在整个协作研发网络中的流动,而知识能否快速有效转移与应用与整个网络的结构有着密切联系。因此,研究协作研发网络的演化及其对技术创新的影响这一问题具有重要的理论与实践意义。已有研究对于协作研发网络演化的研究主要集中在地理集群效应、网络特性分析、影响因素的定性分析方面,缺乏对协作研发网络演化过程的定量描述及其演化对技术创新影响的实证研究。同时,由于协作研发网络本身具有复杂性的特点,本文运用复杂网络理论与分析方法,社会网络分析理论与方法对协作研发网络的形成与演化机理进行深入的理论分析,并定量描述协作研发网络演化的过程；采用基于Agent的建模与仿真方法与计量经济模型对协作研发网络网络结构的演化对技术创新的影响进行模拟仿真与实证检验。主要的内容包括：(1)分析协作研发网络的形成与演化机理。开放式创新背景下,研发边界开始超越企业,合作剩余的获取为协作研发活动存在提供了现实基础。首先,在通过对协作研发网络创新主体之间微观合作机制分析的基础上,探讨了协作研发网络组织的典型特征,包括主体多样性、协同性、动态性、非线性以及嵌入性等特点,并对协作研发网络不同的组织模式与类型进行了对比分析,阐述了协作研发网络所具有的竞争优势。其次,基于生命周期理论分析了协作研发网络的演化过程,并描述了协作研发网络在各个阶段的主要特征。再次,利用基于Agent的建模与仿真方法分析了协作研发网络创新主体之间的跨组织学习过程,指出了知识转移必然导致协作研发网络知识差异的逐步降低。(2)分析协作研发网络的网络特性对技术创新的影响机理。基于协作研发网络具有显著的复杂网络特性,分析了协作研发网络的小世界性、无标度性、网络中心性以及整体密度。对协作研发网络的网络密度、网络中心性以及集聚系数对技术创新的影响进行了理论分析,在理论分析的基础上,提出了协作研发网络的网络密度、网络中心性、集聚系数与技术创新之间存在倒U型关系的研究假设。同时,考虑到不同类型的技术创新对知识需求存在差异,对协作研发网络网络结构演化对突破式创新与渐进式创新的影响进行的对比分析,提出了协作研发网络的网络密度、网络中心性、集聚系数与突破性创新之间存在倒U型关系的研究假设,以及协作研发网络的网络密度、网络中心性、集聚系数显著正向影响渐进式创新的研究假设。(3)利用基于Agent的建模与仿真方法构建协作研发网络的知识增长模型并进行模拟仿真。基于对协作研发网络的复杂适应性分析以及基于Agent建模与仿真技术与复杂适应系统研究之间关系研究,采用基于Agent的建模与仿真技术来构建协作研发网络的知识增长模型。从搜索规则、扩散规则、溢出规则以及创新规则四个方面描述协作研发网络内创新主体的知识行为,构建了协作研发网络的知识增长模型。并采用Netlogo仿真平台模拟不同网络结构情况下协作研发网络的知识增长情况,收集数据,利用回归模型对协作研发网络结构演化与知识增长之间的关系进行了检验。仿真结果从知识管理的视角表明协作研发网络的网络密度、网络中心性以及集聚系数与技术创新之间存在倒U型关系。(4)以汽车产业为例构建协作研发网络,定量描述协作研发网络的演化过程,并对网络结构的演化与技术创新关系进行实证研究。在收集汽车产业联合申请专利数据的基础上,采用Ucinet软件构建1986-2009年历年的协作研发网络,并定量描述了协作研发网络的演化过程。在此基础上收集汽车产业研发投入与产出数据,应用协整理论与方法进行实证研究,即对协作研发网络的网络密度、网络中心性以及集聚系数与技术创新、突破式创新、渐进式创新关系的相关研究假设进行了实证检验。实证结果表明了协作研发网络的网络密度、网络中心性以及集聚系数与突破式创新之间存在倒U型关系,网络密度正向显著的影响渐进式创新。最后,在总结实证研究结论的基础上,结合全文的研究,从国家、产业、企业三个层面提出了推动技术创新能力提升的政策建议与管理对策。更多还原

【Abstract】 With the higher speed products renewal and the fiercer market competition, and the wider demand of technology innovation on professional abilities, the structure of the collaborative relationships among organizations within the industry turns into a close network rather than two points, thus a R&D collaboration network has been occurred. For the sake of the cooperation among organizations and the sharing of technology innovation resources, the knowledge have to flow in the R&D collaboration network, and the key point for quick and efficiency knowledge flow is the structure of the whole network. Therefore, studying on the evolution and impact on technology innovation of the R&D collaboration network is a important issue with theoretical and practical significance.Informed research focus on geography constellation effect, network characteristics and influence factors of the R&D collaboration network in terms of qualitative analysis, but lacking of quantitative description on evolution process and empirical research on the impact on technical innovation. This paper makes up for the deficiency, and meanwhile, using the theory and analysis methods of the complex network and the social network to have a deep theoretical analysis on the formation and evolution mechanism of the collaborative R&D network, and quantitatively describes the evolution process of the R&D collaboration network, and using the Agent modeling and simulation method and the econometric model to simulate and empirically test the impact on technical innovation from the evolution of the structure of the R&D collaboration network. The main contents including:(1)The formation and evolution mechanism of the R&D collaboration network. Under the background of open innovation, The R&D edge begins to cross the enterprise, and the acquisition of cooperative surplus makes the collaborative R&D activities possible. First, based on analyzing the micro cooperative mechanism among agents of the R&D collaboration network, this paper discusses the typical characteristics of the R&D collaboration network organizations, including the diversity, cooperativity, dynamics, nonlinearity and embedability. And comparatively analyzes various organization patterns and types of the R&D collaboration network, and concludes the competitive advantages of the R&D collaboration network. Second, this paper analyzes the evolution process of the R&D collaboration network based on the life circle theory, and describes the leading features of the R&D collaboration network at each stage. Third, this paper analyzes the inter-organization learning process based on the Agent modeling and simulation method, and points out that the knowledge transfer will lead to the decrease of knowledge differences of the R&D collaboration network.(2)The impact mechanism of the network characteristics of the R&D collaboration network on technical innovation. As the R&D collaboration network is a very typical complex network, this paper analyzes the small world characteristics, scaleless property, network centrality and network density of the R&D collaboration network. And then analyzes the impact on technical innovation by the network density, network centrality and cluster coefficient of the R&D collaboration network, and then proposes a research hypothesis that the network density and the network centrality and the cluster coefficient respectively have an inverted U-shaped relationship with technical innovation. Meanwhile, considering the knowledge demand for various technical innovation is different, and the comparative analysis between the impacts on radical innovation and incremental innovation by network structure evolution of the R&D collaboration network, this paper proposes two research hypotheses that the network density and the network centrality and the cluster coefficient respectively have an inverted U-shaped relationship with radical innovation, and that the network density and the network centrality and the cluster coefficient respectively have a significant positive relationship with the incremental innovation.(3)The analog simulation based on the agent modeling and simulation method and the establishment of the knowledge increase model of the R&D collaboration network. After the complex adaptability analysis of the R&D collaboration network and the connection between Agent modeling and simulation techniques and the complex adaptability analysis, this paper adopts Agent modeling and simulation techniques to establish the knowledge increase model. And this paper describes the knowledge behaviors of innovation agents within the R&D collaboration network from four aspects, including search rules, diffusion rules, overflow rules and innovation rules, then to establish the knowledge increase model. And then use the Netlogo simulation platform to simulate the knowledge increase under different network structures. And this paper also collects data and uses the regression model to empirically test the relationship between the evolution of the network structure and the knowledge increase. From the perspective of knowledge management, the simulation results show that the network density, centrality and cluster coefficient of the R&D collaboration network have an inverted U-shaped relationship with technical innovation, respectively.(4)Establish the R&D collaboration network by using the data of the automobile industry, in order to have a quantitative description on the evolution process of the R&D collaboration network, and test empirically on the relationship between the evolution of the network structure and technical innovation. After collecting the data of joint patent of the automobile industry, use Ucinet software to establish the R&D collaboration networks from1986to2009, and have a quantitative description on the evolution process of the R&D collaboration network. After collecting the R&D input and output data, this paper uses the co-integration theory to do the empirical test on the research hypotheses that mentioned above, including whether the network density and the network centrality and the cluster coefficient have a relationship with the technical innovation, or with the radical innovation, or with the incremental innovation, and what kind of relationships they are. The empirical results show that the network density, centrality and cluster coefficient of the R&D collaboration network, respectively, have an inverted U-shaped relationship with the radical innovation, meanwhile, the network density has significant positive effect on the incremental innovation.In the end, based on the empirical research conclusions, this paper combines all the researches, and then proposes policy suggestions and management countermeasures to promote the technical innovation from the perspectives of enterprise, industry as well as country.更多还原