Research Of Urban Road Network Invulnerability Based On Complex Network

With the increase of the urban road traffic volume and the seriousness of the traffic problems, research on the invulnerability of urban road network appears more and more important. It is quite important to improve the work efficiency of urban road network to evaluate the invulnerability of urban road network reasonably and objectively. With the deep-going research on the complex network theory in recent years, it provides a new method for us to analyze urban road network.This paper studies the invulnerability of urban road network based on complex network. This paper firstly introduces the basic theory and the invulnerability theory of complex network. Next, two modelling methods of urban road network and the virtues and defects of the two methods are introduced. According to the basic statistical characteristics of complex network, this paper has proposed the basic statistical characteristics of urban road network and has given the quantitative description. The accuracy of degree-rank function and degree distribution in scale-free networks and exponential networks is analyzed, respectively. According to the comparative analysis of the evaluating indicators of complex network invulnerability, the relative size of the biggest connected subgraph and the potency of the whole network are chosen to be taken as the evaluating indicators of the invulnerability of urban road network, and the quantitative description is carried on.Finally, the road network of the main urban district of Changsha is taken as an example. After carrying on the modelling, the basic statistical characteristics of the network are calculated. Then the fitting analysis of the degree distribution function is carried on. The result shows that Changsha's urban district road network obeys power law distribution, so Changsha's urban district road network is a scale-free network. Then the accuracy of degree-rank function and degree distribution in this scale-free network is studied, and the result indicates that degree-rank function is more accurate than degree distribution. The changes of the relative size of the biggest connected subgraph and the whole network potency under the stochastic attack and the selective attack are analyzed respectively. Then the comparative analysis of the invulnerabilities of the network under different forms of attacks is carried on. The result indicates that the invulnerability under the selective attack is obviously weaker than that under the stochastic attack.