复杂产品系统研发过程风险累积的网络演算度量方法研究

Complex product systems (CoPS) are characterized by complexity, dynamic, high innovation and high risk. Risk comes mainly from research and development (R&D) process. Risk management during R&D process has a significant impact on project success. And risk measurement plays an very important role. The research object of this application is CoPS R&D process risk. Firstly, the information gap theory is introduced into CoPS risk measurement framework, and the interaction mechanism between information gap and risk is revealed. Then the dynamic evolution pattern and the coupling relationship of R&D process risk are analyzed, and a double-layer network mapping model between information and risk is constructed. Finally, according to network calculus theory，risk accumulation deterministic boundary measurement during R&D process is realized by establishing risk information flow model and service model. Through describing risk with information gap, the implicit risk can be represented, which could overcome the difficulty that risk factors can’t be acquired directly due to the exploration feature of CoPS, and could reduce cumulative measure errors from existing methods. At the meantime, the application also develops a worst case deterministic boundary measurement for risk based on network calculus, which could evaluate accumulative risk effect. The achievements generated in the research could lay an important foundation of risk management technology and method for improving CoPS risk control level, and also could enrich and develop risk management theory.

复杂产品系统具有复杂动态、高创新性、高不确定性的特征，风险主要来源于研发过程，具有风险累积和演化特征。研发过程风险管理对项目成败影响重大，其中的关键问题在于风险度量。本项目以复杂产品系统研发过程风险为研究对象，将信息差纳入风险度量框架中，揭示信息差与风险的作用机理；分析研发过程风险的耦合关系，提取风险的动态演化规律和累积特征，构建信息与风险的双层网络映射模型；建立风险信息流模型和服务模型，研究网络演算视角下研发过程风险累积的动态确界度量。本项目通过信息差对风险进行描述，使很难直接识别的风险能够显性表征，可以克服由于复杂产品系统的探索性导致风险因素难以直接获取的困难，以及在此基础上度量造成的累积误差；通过网络演算建立风险最坏确界量化方法，实现对具有风险累积特征的复杂产品系统研发过程风险的度量。项目研究成果能丰富和发展风险管理的理论，为复杂产品系统风险控制水平的提高提供技术和方法的重要支撑。

复杂产品系统（CoPS）规模巨大、结构复杂，具有高度的探索性和创新性，跨组织的多主体高度参与，组成系统的各个要素之间存在着庞杂的相互作用，而且风险贯穿于CoPS研发过程，在研发过程的各个阶段动态演变，导致风险可靠度量的困难。复杂产品系统涉及的领域比较广泛，并不局限在一个行业或企业。比如：航空航天系统、大型武器装备、大型船只、高速列车、智能交通系统、大型电信通讯系统、电力网络控制系统等，直接影响到一个国家的综合国力和核心竞争力。本项目从理论建模、算法设计、模拟仿真、案例分析等多个角度开展研究，并通过在不同领域的深入调研，以飞机、航空电子综合系统、导弹综合电子系统等大型、复杂、集成度高的系统为研究切入点，去探索其本身的特征和风险规律，从信息和复杂网络的视角抽象出研究对象的本质，基于信息差和网络演算，提出相应的风险识别、传导和度量方法，并通过仿真或实例对模型进行了验证。通过本项目研究，实现了基于信息去表征CoPS风险，基于网络去刻画风险关联和风险传导，解决了量化CoPS研发过程风险的难题，拓展了风险管理的视角，实现了从不同角度去测度风险的可能性。相关研究成果为复杂产品系统风险管理奠定了理论基础和实践指导。