Network Model And Research On Invulnerability For Transportation System

Transport system is the prerequisite for economic development, its stability andreliability directly influence the normal development of the social economy. Along with theresearch of the complex network more lucubrate, the theoretical significance and applicationvalue of invulnerability research in transportation network based on complex network becomeincreasingly important. This paper studies the survivability analysis and optimization of thetransport network by combined transport network self-characteristics and the knowledge ofgraph theory, probability theory, statistical physics, particle swarm optimization algorithm,computer simulation, and other multi-disciplinary domain. The research works includeevaluation of node importance in transport network, optimization of connectedness androbustness of the railway network based on adding edges, optimization of connectedness ofload-capacity nonlinear model in weighted network for cascading failure, robustness analysisof transport network against cascading failure. Research and innovation are as follows:(1) Transport network is a weighted network which is formation of transportation nodeand the connection between nodes. The evaluation of node-importance has very strongguiding significance for the systems management, disaster prevention, road planning, andoperation scheme designing.We proposed the weighted interdependent network model of transportation system toevaluate the importance of node in transport network. In terms of the node importance equalto the conspicuousness of the node for the connection between it and others, the ResourceAllocation Algorithm (RAA) is proposed in this paper. The algorithm could calntiemtwe ocrokm vpculate theresource allocation between stations, then obtains the importance value of stations, and itselreifxyit yth eis f e a s(i4b i l i2ty)a.n Td heef fescimtivuelanteiossn so fb tahsiesd m oenth othde, aranidlw ita cyo nuledtw eovraklu aanted thseu bnwodayeimportance of transport network.(2) Transport system plays an important role in the development of the national economyto assume the economic ties and personnel carriers, and its survivability is important for highadaptation, emergency and security function.We describe and evaluate the passenger rail transport system based on the weightedrailway traffic network and the algebraic connectivity, respectively. We optimize theconnectedness and robustness of the passenger rail transport system by maximizing thealgebraic connectivity of the weighted railway traffic network based on adding train fluxes (edge-weight). A Discrete Particle Swarm Optimization based on Sorting Strategy andDirected Chaos-mutation (SSDC-DPSO) is proposed in this paper for the optimizationproblem. Subsequently, comparing with the Modified Greedy Perturbation (MGP) algorithmand the Weighted Tabu Search (WTS), the advantage and effectiveness of SSDC-DPSOalgorithm was proved in solving the optimization problem.(3) The transport system is a comprehensive, dynamic, open complex giant system. Theinvulnerability optimizing strategy based on dynamic process is important for theinvulnerability optimizing of transport system.In the study of robustness of complex network against cascading failure, reasonableallocation of capacity could achieve strong robustness with less cost. We study the cascadingfailure based on the load-capacity nonlinear model. For the parameters optimization problemin the model, confirming the ranges of parameters by theoretical analysis, optimizingparameters by the Multi Objective Particle Swarm Optimization (MOPSO). The simulationproved that the proposed optimization method has better control force to cascading failure, itcould reduce network costs in the case of strong robustness.(4) The invulnerability analysis based on dynamic and topological structure is importantfor the planning, design, and maintenance of transport system.We investigate the cascading failure on weighted complex networks by focusing on therelation between capacity and load, and the Load-Capacity Optimal Relationship (LCOR)model is proposed to improve the network robustness against cascading failure with less costin this paper. By comparing with other three kinds of load-capacity linear or non-linearrelationship models in model networks and real-world weighted networks, we prove that theLCOR model has high robustness against cascading failure with the least cost. In order toprove the effectiveness and universal of the LCOR model, we provide some theoreticalanalysis and give the computational method of its cost threshold. The simulations on railwaynetwork, the airports network and the metro network prove that our findings have greatapplicability for the design of more robust real-world networks with less cost to againstcascading failure.