Research On Optimization Of Complex Network Structural Controllability And Attack Robustness

In recent years,the research of complex networks has developed rapidly.Complex network theory and related technologies have become important tools and methods for describing the structure and dynamics of complex systems.In particular,the research on the robustness and controllability of real systems in real life has become a hot topic in the field of complex networks.Based on the theory of structural controllability and network attack,combined with robust optimization method,this paper mainly studies the changes of the structure and performance of real networks after the improvement of anti-attack ability.The relationship between the anti-attack ability and the controllable ability of the real networks is analyzed in detail,and the data mining and analysis software of the complex system based on network models is designed and developed.The main contents and innovative results of the paper are summarized as follows:(1)Analysis of the structure and performance of the network after anti-attack capability optimizationBased on the optimization algorithm of preserving degree invariant and edge crossing reconnection,this paper firstly investigates the structure changes of European aviation network,European power network and American aviation network.Through visualization of the network topology connection structure before and after optimization,it is found that the optimized network presents an Onion-like shape.At the same time,the variation law of important characteristic parameters of the network in the optimization process is deeply studied.Then,from the network hierarchy angle,the edge connection probability of nodes with the same degree and the change rule of the proportion of nodes with different k-shell values are analyzed,thus we have a deeper understanding of the topology properties of the network before and after optimization.Finally,the optimization effect of real network anti-attack ability and the change of network synchronization ability before and after optimization are investigated.(2)Research on controllability of large-scale networks with Onion-like structureFirstly,ten classic real networks are taken as research objects,and the numerical control method is adopted to analyze the change of controllability in the real network anti-attack ability optimization process.The results show that the robustness and controllability of the optimized real network are enhanced.Then the robustness and controllability of scale-free networks are compared and studied.Simulation results show that increasing the average degree and power exponent of scale-free network can improve the controllability of network simultaneously,and the controllability is also improved in the process of robustness optimization.(3)Design and implementation of data mining and analysis software for complex systems based on network modelsA simulation software is designed for the study of large-scale complex systems.By means of numerical simulation and statistical analysis technology,the software completes the structural characteristics analysis,modeling,dynamic behavior simulation,comparative analysis,analysis results display and other functions of large-scale network systems.The software adopts a module-based hierarchical design to ensure that the software structure is clear and readable.Modules can be reused and maximized for reliability,stability,scalability and ease of maintenance.