【作者】 王智生；

【导师】 胡珑瑛；

【作者基本信息】 哈尔滨工业大学 ， 技术经济及管理， 2013， 博士

【摘要】 随着经济全球一体化进程的加深，作为市场主体的企业所面临的全球化竞争日益激烈，而面对科学技术的飞速发展及对企业创新能力要求的不断提高，单一企业受困于研发成本的不断攀升而加剧了运营风险。在这样的背景下，研发联盟因其在攻克关键共性技术、拓展产业市场空间、服务专项任务目标、促进产业良性发展、参与国际合作竞争等方面显现出来的优势而蓬勃兴起。然而，由于研发联盟受到资源投入大、技术复杂、协同创新风险等不确定因素的影响，成员间的信任关系作为合作的基础和联盟的重要治理机制至关重要。深入分析和揭示研发联盟成员信任关系的动态演化机理以求解决联盟治理是研发联盟技术创新面对的重要问题。在对国内外相关研究成果系统梳理和总结的基础上，运用博弈论、系统科学理论等规范研究方法，与实际调研、机器学习等实证研究方法相结合，从知识转移视角系统深入的研究了研发联盟成员间信任演化的内涵、过程、动力以及演化中的信任评价和修复问题。基于知识转移视角，揭示研发联盟信任演化机理，主要从以下几点展开并得出结论：首先，在对前人的组织信任及信任演化理论的研究成果进行系统梳理与归纳总结的基础上，界定了研发联盟信任及信任演化的内涵，分析了联盟信任的影响因素，借鉴并采用计算型信任、知识型信任、认同型信任三层次的经典信任演化研究思路，并将研发联盟成员间转移的知识分为技术知识、系统知识以及战略知识三类进行分析，进而从知识转移视角构建研发联盟信任演化研究的体系结构，为研发联盟信任关系理论研究与实践提供了一个崭新的视角和切入点。其次，基于生命周期理论将知识转移类型、信任演化过程与联盟形成、发展及成熟三个阶段相结合，构建基于知识转移的研发联盟信任演化机理的理论模型。通过对联盟、知识转移以及信任的相关文献的分析，提出基于知识转移的研发联盟信任演化作用机理的相关研究假设，以技术知识、系统知识、战略知识转移作为自变量，计算型信任、知识型信任、认同型信任作为因变量，并引入经典的组织信任整合模型中信任关键影响因素——（感知到的）能力、诚实与善意作为中间变量（其中感知到的能力与诚实属于信任知识方面的因素，善意代表可信性情感方面的因素），并将联盟的三个发展阶段的知识转移与信任演化理论分析结合，建立了一个贯穿联盟三阶段的信任演化整合理论模型。然后，在理论分析的基础上，深入实际研发联盟访谈、发放问卷进行实际调研，实证检验了基于知识转移的研发联盟信任演化机理。结果表明，技术知识转移与（感知到的）能力正相关，战略知识转移与诚实、善意正相关，而系统知识转移对能力、诚实、善意均正相关；能力、诚实与计算型信任正相关，善意与认同型信任正相关，而三者与知识型信任均正相关；在联盟的形成、发展及成熟阶段，计算型、知识型、认同型信任分别对技术知识、系统知识与战略知识转移具有正向促进作用。再次，从知识转移视角分析研究了研发联盟成员间信任演化的动力机制，运用系统科学研究方法，分别构建了基于知识转移的信任演化动力学模型与含时滞动力学模型。求解信任演化动力学模型并绘制其相位图，定义三个区域分别代表创新发展、动荡以及崩溃区域，并提出三项促进联盟信任向创新发展区域演化的对策和措施；对含时滞信任演化动力学模型求解得到，在一定条件下(a≤b)，信任函数与知识转移函数之间一定会发生周期变化现象。最后，分析了基于知识转移的研发联盟信任演化路径，在此基础上，对演化过程中的信任评价与信任修复问题进行深入分析和研究。通过对信任不信任二维性与信任非对称性分析，提出了基于知识转移的信任演化路径。运用粗糙集与小波支持向量机的分析研究方法，构建了基于知识转移的信任评价模型，实证结果验证了信任评价模型的有效性及精度。将信任研究对象分为研发联盟的领导者与一般成员两类，以研发联盟领导者对联盟成员在知识转移过程中发生信任违背后采取信任修复决策的问题为背景，构建了基于知识转移的信任修复羊群行为模型；结果表明，拥有私人信息的联盟领导者在有限理性的情形下可能发生羊群行为，并通过算例对信任修复羊群行为模型及发生条件进行了验证。更多还原

【Abstract】 With the further deepen of economic globalization, as a market subject,companies are facing global competition increasingly fierce. With the rapiddevelopment of science and technology and the increasing requirement ofinnovation capability, the operational risk of a single enterprise trapped in R&Dcosts will intensify. Due ot the enormous advantage of capturing key generictechnology, expanding industrial market space, providing services for specificobjectives, promoting the benign devlopment of industry, participating ininternational cooperation and competition and other aspects, R&D alliances aregrowing rapidly in this context. However, because of large resource investment,technological complexity, collaborative innovation risk, trust relationship betweenR&D alliances members, as the basis of cooperation and alliances importantgovernance mechanism, is crucial. The depth analysis of the dynamic evolutionmechanism of trust relationship between R&D alliance members, in order to solvethe alliance governance, is the key issues to R&D alliance carrying on technologicalinnovation.Based on collating and summarizing relevant research results at home andabroad systematically, this paper which makes a combination use of norms scientificresearch methods, such as game theory and system science theory, and empiricalmethods, such as the actual research and machine learning, from the perspective ofknowledge transfer, makes intensive research on the connotation, process, dynamicsof trust evolution and trust evaluation and repair between alliance members. Thispaper reveals the mechanism of trust evolution between alliance members based onknowledge transfer, and commences from the following points and drawsconclusions as follows:Firstly，by analysing theoritical research results of organizational trust and trustevolution, using three level of classic trust evolution model structured bycalculus-based trust, knowledge-based trust, and identitication-based trust forresearch ideas, this paper defines the connotation of R&D alliance trust and trustevolution, analyzes the factors of influencing alliance trust and types of knowledgetransfer divided into technological, systemic and strategic knowledge betweenalliance members, and constructs an architecture of trust evolution in R&D alliancefrom the perspective of knowledge transfer. That provides a new perspective andentry point for trust theory study and practice of R&D alliance.Secondly, on account of life cycle theory, this paper which types of knowledgetransfer, process of trust evolution, and three stages of alliance formation, development and maturity combined, construct a theoretical model that revealingthe mechanism of trust evolution in R&D alliance based on knowledge tranfer.Analyzing the relevant literature of alliance, knowledge tranfer and trust, this paperproposes hypotheses about the trust evolution mechanism of R&D alliance based onknowledge transfer, combines theoretical analysis of knowledge tranfer and trustevolution throughout the three stages of alliance developing, and constructs anintegrated theoretical model of trust evolution throughout the three stages of alliance,which related to technological knowledge, systemic knowledge, strategic knowledgetransfer as independent variables, calculus-based trust, knowledge-based trust,identitication-based trust as the dependent variable, and key factors as intermediatevariable introduced from a classic integrated model of organizational trust.Then, based on theoretical analysis, this paper makes an empirical test on themechanism of trust evolution in R&D alliance by in-depth interviews andquestionnaires for the actual research. The results indicate that technologicalknowledge transfer has a positive effect on (perceived) ability, whereas strategicknowledge transfer is positively related to integrity and benevolence, and systemicknowledge transfer is positively correlated with ability, integrity and benevolence.The results also demonstrate that ability and integrity are positively correlated withcalculus-based and knowledge-based trust, whereas benevolence has a positvieeffect on identitication-based trust. At the formation, development and maturationstage of alliance, calculus-based, knowledge-based trust, and identitication-basedtrust is positively correlated with technological, systemic and strategic knowledgetransfer separately.Again, with the perspective of knowledge transfer, the dynamic mechanism oftrust evolution between R&D alliance members is analysed. And with the method ofsystem science, a dynamics model and a delayed dynamic model of trust evolutionbased on knowledge transfer were constructed. This paper solves the dynamicsmodel of trust evolutionary, draws the phase diagram, defines three areasrepresenting the developing, unstable and collapse region of collaborativeinnovation, and proposes three measures promoting trust to the developing region ofcollaborative innovation. Solving delayed dynamic model of trust evolution, thispaper founds that, under certain conditions (a≤b), periodic variation of trust andknowledge transfer functions will happen.Finally, on the basis of analyzing trust evolutionary path in R&D alliance fromknowledge tranfer perspective, this paper mainly explores trust evaluation and trustrepair in the process of trust evolution. Analyzing the two-dimensionality of trustand distrust and trust asymmetry, trust evolutionary path is proposed on the basis ofknowledge transfer. Applying rough sets and wavelet support vector machinemethod, trust evaluation model is bulid and the empirical results verify the validity and accuracy of it. In the context that the subjects were divided into leaders andgeneral members, and the leaders in R&D alliance make trust repair decisions whentrust violation occur, this paper constructs herd behavior model of trust repair basedon knowledge transfer. The results demonstrates that, herd behavior may occur whenalliance leaders who have private information in the case of bounded rationality. Inthe end, herd behavior model of trust repair and its occurred condition are verifiedthrough numerical examples.更多还原