基于微分动力学方程的相互关联网络故障传播联合演化模型

Joint Evolution Model of Fault Propagation in Coupled Networks based on Differential Dynamics Theory

研究网络遭受相继故障后的演化机理是网络科学的重要课题之一。现有研究少有涉及到关联网络间的相互影响作用。随着网络中节点复杂度和智能化的提升,仅考虑单个网络节点之间的耦合关系是难以客观反映网络状态的演变规律。针对现有渗流模型只便于分析异构网络之间一一对应关系的情况,本文中采用弱化网络之间作用形式而强调作用效果的方法,建立了有电网供能的计算机网络相继故障传播演化模型。同时通过将电网和计算机网络构成的庞大的联合网络投影到计算机网络层面(即只保留至少一端为计算机网络节点的连边),降低了模型的难度。通过模拟实验,电站小规模的随机故障可引发计算机网络相继故障,并达到使其发生大规模随机故障的效果。

One of the most important topics in network science is to explore the evolution mechanism of networks after successive failures. And few studies involve the interaction between interconnected networks. With the promotion of complexity and intelligence of nodes, it is difficult to objectively reflect the evolution law of the actual network state only considering the coupling relationship of between single network nodes. Based on mutually beneficial symbiosis, a joint evolution model of related networks is established. Because the existing seepage model is only easy to analyze the one-to-one correspondence between heterogeneous networks, a method to weaken the form of interaction between two networks but emphasize its effect is proposed, while the evolution model of the computer network based on the power network is established. At the same time, the difficulty of the model is reduced by projecting the huge joint network of power grid and computer network to the level of computer network. Simulation experiments indicate that the small-scale random failures of power stations may cause successive computer network faults and achieve the effect of large-scale random faults.