作者: Dominik Heide , Mirko Schäfer , Martin Greiner
DOI: 10.1103/PHYSREVE.77.056103
关键词: Robustness (computer science) 、 Topology 、 Network structure 、 Mathematical optimization 、 Limit (music) 、 Complex network 、 Mathematics 、 Stochastic process 、 Cascading failure
摘要: Fluctuating fluxes on a complex network lead to load fluctuations at the vertices, which may cause them become overloaded and induce cascading failure. A characterization of one-point is presented, revealing their dependence nature flux underlying structure. Based these findings, an alternate robustness layout proposed. Taking correlations between vertices into account, analytical prediction probability for remain fully efficient confirmed by simulations. Compared previously proposed mean-flux layouts, comes with significantly less investment costs in high-confidence limit.