倒向协同创新项目治理网络研究

How to improve the controllability of collaborative innovation projects is a difficult problem. It cannot be solved by solely relying on science and technology leaders while ignoring the impacts of project governance. Neither can it be solved by forward project governance methods based on predetermined stakeholder boundaries and related static network structures. To solve this problem, this research proposes to apply a backward research method. That is, based on certain risk thresholds set by governance subjects, the stakeholder boundaries are defined backward and related dynamic relationships of project governance networks are constructed in order to reduce risks in collaborative innovation projects gradually. To solve the contradiction between the uniqueness of a project and the generalizability of research methods, this research proposes to classify risks into two types, namely attribute risks and structure risks, treat the stakeholders’ attributes as aspects of the nodes of the project governance networks, simplify project governance activities as relations between information channels and treat them as aspects of connection lines between nodes, define governance behaviors as utility functions of network node adjustments and information channel relationships between the nodes, and determine the values of project structure risks by measuring the entropies of social networks formed by project stakeholders. Based on certain risk threshold objectives, the stakeholder boundaries of collaborative innovation projects, the governance network structures and their risk values can be determined iteratively by a dynamic evolutionary process of project governance networks, which integrates the methods of social network analysis and system dynamics. This research will provide new theories and methods to solve the problem of controllability in collaborative innovation projects.

如何提高协同创新项目的可控性是个难题，仅依赖科技领军人物而忽视项目治理的作用不能解决它，基于预先设定的利益相关方边界及其静态结构的正向项目治理也不能解决它。本课题将采取倒向研究的方式，根据治理主体设定的风险阈值，倒向确定利益相关方边界并构造相应的项目治理网络动力关系，逐步降低协同创新项目的风险。为解决项目的独特性和研究方法的统一性之间的矛盾，本课题将项目风险归结为属性和结构风险两种类型，将利益相关方属性纳入项目治理网络节点，将项目治理活动简化为信道关系并纳入点间连线，将治理行为表达为调整网络节点的效用函数及节点间信道关系两种方式，项目结构风险值将通过项目利益相关方社会网络熵的评价确定。根据给定的风险阈值目标，协同创新项目的利益相关方边界、治理网络结构及其风险值将通过集成了社会网络分析和系统动力学方法的项目治理网络动力演化过程迭代得出。本课题将为解决协同创新项目可控性问题提供新的理论和方法。

如何提高协同创新项目的可控性是个难题，基于预先设定的利益相关方边界及其静态结构的正向项目治理不能解决它。本课题提出了“倒向”风险管控思想，即根据治理主体设定的风险阈值，倒向确定利益相关方及其关系构成的项目治理网络，进而提出了网络动力分析（NDA）方法。.根据课题负责人上一国家自然基金项目成果之一——项目治理概念，从结合NDA的动态和网络两个特点出发细化项目治理含义。同时，从更能体现管理学价值和项目特点的“人”的角度，将项目治理风险定义为“利益相关方不确定行为的可能性、影响程度和可管理程度”；并根据项目的独特性和临时性特点将利益相关方行为界定为“围绕项目风险因素管控而开展的系列活动”。.基于上述概念，通过利益相关方行为影响因素分析，将项目治理网络关系分为沟通和控制两类。并据此进一步提出网络节点界定规则：内部信息传递有效、内部目标统一的利益相关方视为网络节点。然后，根据风险分解结构、利益相关方风险角色来识别网络节点；通过分析节点沟通和控制两类关系以识别网络关系；设计基于最优最劣方法的关系强度判定过程，以量化治理网络。.通过分析项目治理网络、利益相关方行为、风险因素间的影响关系来探究项目治理网络动力过程机理。通过网络密度、节点网络距离等指标量化沟通网络和控制网络，并建立了上述指标与利益相关方行为评价指标间的数理关系。根据该机理，设计网络动力过程算法结构，并在MATLAB上进行程序实现。.基于“倒向”风险管控思想，以NDA为核心，设计了项目治理风险管控决策的流程。然后，选择治理风险较高的工业技术研究院协同创新项目作为案例，展示NDA的应用过程。并基于案例应用过程，结合专家意见对NDA进行案例验证；采用实证研究方法对NDA的相关结论进行统计验证；借鉴系统动力学模型验证方法，对NDA进行仿真验证。.最终，本课题提出了一套能够根据风险阈值目标倒向分析治理网络结构演化过程、并逐步降低项目治理网络风险，提高协同创新项目可控性的理论和方法。