| The complex networks are becoming increasingly perfect that include the physics,computer science,biology,etc.The powerful function of infrastructure system takes a great convenience to people's life.Related research has shown that,infrastructure networks play an important role in real life,such as power grids,airline routes,World Wide Web and transportation networks.Internal dynamics characteristics become ever more obvious,with the internal relevance increasing.A targeted attack or random failure in a network may cause cascading failure,which leads to serious consequences.Currently,preventing and controlling the cascading failure has become the front and hotspot in the area of infrastructure systems security in domestic and overseas.However,there are still existing some shortcomings on the aspect of the influence of simple networks on cascading failure.Firstly,the majority of research results related to the cascade model have concentrated on the single index,which to control and prevent cascading failures,but overlooked the facts that many cascading failure is related to the mixing index closely in infrastructure networks.Secondly,there is still no researcher studied the models with different loads under the same network conditions.Thirdly,there is no visual system to let the researchers who study the robustness of power grids with cascading failure in complex networks.This made the threshold of scientific research has been raised and the difficulty of research has been increased,which need us to do a future research.To solve these problems,this thesis intends to study on the critical dynamic of redistribution loads strategy and studies cascading failure models under different loads.Also,we propose the optimization strategies and develop software for power grids robustness research.The new software visualizes the simulation operation of the complex network and improves scientific research efficiency.The main contents are as follows:1.The classical theories and methods of complex networks are studied.Introduce the main topological models and algorithms of complex networks.Theoretical analyze the process of cascading failure with traditional models.Five universal robustness metrics is proposed.2.Cascade models suitable for different loads are proposed.Considering how to more effectively protect the network to avoid the global collapse due to cascading failure,the new models with the betweenness centrality and the degrees of the nodes which are combined.The principle of local optimal redistribution is adopted to simulate the process of cascading failure.The cascading propagation is analyzed by increasing the capacity of every edge.The effects of the parameters of the edge weight on cascading dynamics are investigated.The simulation results show that the combination of the betweenness centrality and the degree of the nodes is a great significance on resisting dynamics process of cascading failure.3.A deduction system for cascading failures in power grids is developed.Taking the.NET as a platform for building the deduction system.The simple,modern,object-oriented,and type-safe programming language C# is used.The deduction system gets the simulation results quickly through comparison,which lowers the threshold and decreases the difficulty of scientific research.To sum up,this thesis put forward two new models with the betweenness centrality and the degrees of the nodes which are combined.The ability to resist cascading failure is studied with the new models.At the same time,the research break through the limitation of previous work on cascading failure from the point of view of dynamics.The real reason caused the cascading failure is analyzed.This thesis also proposed effective strategies to control and prevent cascading failures in the BA network.Simulation software for cascading failure is developed.We numerically try to find the better cascade models,which could provide scientific evidence and reliable guidance for maintaining and developing infrastructure networks... |
PDF Full Text Request