Research On Cascading Failure Model And Recovery Strategy Of Complex Networks

As the real-world network system’s overall efficiency continues to improve,the dynamic characteristics within the network are becoming increasingly complex,and network interconnectivity is on the rise.These networks can be viewed as complex systems for research purposes.However,complex networks are vulnerable to various internal and external threats,causing them to become fragile.Even a single malfunctioning component can trigger a chain reaction of failures due to coupling relationships between nodes,leading to cascading failures.Cascading failures can cause substantial damage to the system and business interruption,ultimately leading to significant economic losses.Therefore,this article aims to investigate complex networks,focusing on modeling methods and recovery strategies for cascading failures.The following section details our specific research objectives.Firstly,unlike the previous static load-capacity model based on topology,the random walk theory is utilized to model node dynamic load and corresponding capacity.Moreover,the biased random walk theory based on betweenness centrality is utilized to develop a cascading failure model that accounts for the non-random transmission of load.Using this model,the effects of the extreme value of instantaneous fluctuating load on cascading failure and complex network robustness are explored.The relationship between node overload probability and the node betweenness at the initial time step is studied by simulation,and the effects of different parameters on complex network resistance to cascading failures are analyzed.Secondly,building upon the above cascading failure model,a resilience-based recovery optimization model and a sequential recovery strategy for three typical types of complex networks are proposed.In this optimization model,cascading failure and recovery happen synchronously,taking into account differences in recovery times.Due to limited recovery resources,multiple indicators are used to calculate node importance values,which are sorted using the Copeland scoring method to achieve sequential recovery of failed nodes.Through simulation research,the influence of multiple parameters on the recovery effect was studied and compared with other methods to demonstrate the advantages of this recovery strategy.Finally,we also apply the cascading failure model to the problem of congestion in urban road traffic network,and propose a recovery strategy for local load redistribution to more efficiently redistribute congestion load.Simulation analysis of network status,network efficiency,and travel time is conducted on the BA scale-free network,part of the real road network in Baoding City,Hebei Province,and the Sioux Falls network.The simulation results confirm the effectiveness of this method in recovering from cascading failures caused by traffic congestion.