关键词: 项目组合, 交互风险, 直觉模糊生态聚类算法, Lotka-Volterra方程

Abstract: 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.

Key words: project portfolio, interactive risk, ecological intuitionistic fuzzy clustering algorithm, Lotka-Volterra model