网络体系效能评估改进FDNA模型

网络体系(system of system, SoS)的效能评估是SoS建设和分析的重点问题。传统功能依赖网络分析(functional dependency network analysis, FDNA)方法可以展示SoS“松散耦合”特性, 但缺乏对组件系统的运行独立性、效能衰减性、拓扑规律性等特征的关注。针对此, 在考虑多态问题的基础上, 通过Markov过程分析, 推导组件系统的自主效能衰减函数与系数, 改进了传统方法中的固定参数, 求得节点的动态效能值。在界定“相关SoS”概念的基础上, 计算组件系统可靠度的重要度, 识别关键节点, 构建网络SoS效能评估函数。以五节点航天SoS为例, 演示评估过程并验证了方法的可行性。

Effectiveness evaluation of network SoS based on improved FDNA model

The effectiveness evaluation of network system of system (SoS) is the key issue of SoS construction and analysis. The traditional functional dependency network analysis (FDNA) method can show the "loose coupling" characteristics of the SoS, but it lacks attention to the operation independence, effectiveness attenuation, topology regularity and other characteristics of the component systems. In view of this, on the basis of considering the polymorphism problem, the self-effectiveness attenuation function and coefficient of component systems are defined through Markov process analysis, the fixed parameters in the traditional method are improved, and the dynamic effectiveness value of nodes is obtained. On the basis of defining the concept of the "relevant SoS", the importance of component system reliability is calculated, the key nodes are identified, and the network SoS effectiveness function is constructed. Taking the five-nodes aerospace SoS as an example, the effectiveness evaluation process is demonstrated and the feasibility of the method is verified.