Modeling two-stage failure mechanism of cascading in cyber-physical power systems

Interdependency in cyber-physical power systems enables efficient monitor and control, but also brings out many threats, leading to catastrophic blackouts. To address this problem, the propagation mechanism of cascading failure in systems is analyzed. In this paper, we propose the partial random coupling systems model, and detail the interactive mechanism between physical power grid and communication network. Then, the modified failure mechanism is present, including two-stage redistribution types of traffic loads in physical power grid, and stochastic routing strategies in communication network. In simulation, the impact of attack scenes and topology structure on systems robustness is studied. Compared with random attack, high degree/betweenness attack results in first-order transition of cascading failure at a critical point, and causes more serious damage on systems. Besides, we have proved the positive correlation between clustering coefficient and robustness. To random attack, the systems consisted of double-star network behave more robust than systems consisted of mesh network, and also indicate the misproportion of increase between average dependent degree and systems robustness.