The Research Of Cascading Failures Dynamic Behavior On Interdependent Networks

In the past decades,network sciences have developed rapidly and become a great and powerful tool for the research on complex systems.However most of previous research focused on the inner relationships within one network,but ignored the interactions between networks.In fact,networks usually interact with each other.Therefore,interdependent network,multi-player network and network of networks can model real systems better and have become the hot topic in the field of complex networks since their concepts were raised in 2010.The current research results revealed that interdependent networks are more fragile than traditional single network.Even the failure of one single node may lead to cascade failures and furthermore the collapse of the whole network.This will create catastrophe for systems.Currently,the mechanics of cascade failures on interdependent networks has not been revealed clearly and completely.On the one hand,interdependent networks whose topology is locally tree-like and not spatially embedded have been greatly researched.On the other hand,most research on spatially interdependent networks still focus randomly interdependent systems because there are few theoretical tools for the analysis of spatially interdependent networks and its topology are not locally tree-like.Furthermore,the extreme vulnerability of interdependent networks put forward new challenges and demands for robustness enhancement strategies.On the one hand,as there are higher risks of catastrophic hazard for interdependent networks,interdependent networks demand robustness enhancement strategies to protect more urgently.On the other hand,the topology of interdependent is more complex and traditional strategies are no longer useful for them so that the corresponding robustness enhancement strategies confront with more challenges.In order to solve the problems above,we proposed a model to describe the interdependent networks with controlled degree of randomness for dependency maps.What's more,we explored the mechanics of cascade failure dynamics of spatially interdependent networks and presented a new strategy for robustness enhancement.The main work of this paper is as follows:1.In this paper,we explored how randomness for dependency maps will influence the cascade failure dynamics of spatially interdependent networks.Firstly,we constructed a model for spatially interdependent networks with controlled randomness for dependency maps.Based on this model,we investigated the relationship between the randomness for dependency maps with the percolation phase property and percolation threshold.Furthermore,we proposed an analysis tool for system's transient property based on respond time at percolation threshold and analyzed cascade failure dynamic property of interdependent networks from another dimension.From simulation results,we found that the randomness for dependency maps is a critical factor for the cascade failure dynamic on spatially interdependent networks.2.We proposed a new robustness enhancement strategy for interdependent networks.First,we compared the existing robustness enhancement strategies.And we analyzed advantage and disadvantages of autonomous nodes selection strategies based key nodes.Finally,we proposed a new autonomous nodes selection strategy based on interactions optimization.Compared with exiting research results,this strategy provided greater robustness imprudent.