项目组合交互风险的网络聚类及协同演化研究

Research of network clustering and co-evolution of project portfolio interactive risk

将交互关系纳入项目组合风险研究框架,分析项目组合交互风险网络复杂性及其演化规律。从交互关系分类入手,将项目组合表征为由多个子系统相互作用构成的交互耦合网络,并提出了基于直觉模糊生态聚类算法的项目组合交互风险网络聚类识别模型;基于扩展的Lotka-Volterra方程建立项目组合交互风险协同演化模型,对比研究具有竞争、共生和偏利偏害效应的项目组合交互风险演化规律。研究表明,项目组合风险效率最大化及系统稳定性的关键在于项目组合交互关系及其强度,同时合理控制组合配置比例。

The system complexity and co-evolution of Project Portfolio Interactive Risk（PPIR）are analyzed by introducing project interdependencies into the research framework of project portfolio risk. On the basis of the classification of the project interdependencies, the project portfolio is characterized to be an interactive coupling network with the complex interaction of multi-subsystems, and an intuitionistic fuzzy and ecological based clustering algorithm is put forward as a method for the network clustering analysis of the PPIR system. A co-evolution model of the PPIR is generated based on an extended Lotka-Volterra model, comparing different mechanism of interactive types of competition, symbiosis and commensalism/amensalism on PPIR. The result shows that the optimization of project portfolio risk efficiency and its system stability will be achieved by optimizing portfolio interdependencies and allocation proportion.