基于风险传播网络和K-shell方法的复杂研发项目风险评价

Risk Analysis of Complex R&D Projects Based on Risk Propagation Network and K-shell Method

研发项目中存在大量相互影响的风险因素,如何准确地量化分析这些风险因素,是开展项目风险管控的重要依据。本文采用设计结构矩阵(DSM)和复杂网络的节点重要度排序方法,提出了改进的K-shell法进行风险评估,以更好地反映复杂风险网络特征对风险因素排序的影响。首先,本文采用“风险事件-风险因素”多领域矩阵(MDM),通过风险事件间依赖关系和风险事件-风险因素间的依赖关系,建立了推导风险因素间关系的模型,并构建了集成的风险传播概率计算模型;然后,综合风险网络中节点的出度/入度和节点间的风险传播概率等因素,构建基于K-shell法的风险网络节点重要性排序模型,并提出了改进的K-shell分解法。最后,通过案例,对某研发项目的风险因素进行排序并识别关键风险因素,验证了本文提出模型和方法的有效性。

There are many interdependent risk factors in the R&D projects.How to evaluate accurately these risk factors is an import basis for performing project risk monitoring and control.Using the Design Structure Matrix(DSM)and the method of ranking nodes in complex networks,this paper presents an improved K-shell method to perform quantitative risk analysis which can reflect the influence of complex risk network.First,we use the"risk event-risk factor"Multi-Domain Matrix(MDM)to deduce the dependencies among the risk factors,and build a combined propagation probability model.Further,by integrating the node's out-degree and in-degree centrality,the risk value of risk factors and propagation probability,we build the K-shell model for ranking nodes in complex risk networks.Finally,an industrial example is provided to illustrate the proposed model.