Research On Reliability Of Urban Road Network Based On Theory Of Complex Network

The urban modernization is speeding up increasingly, more and more people are flocking to the center of city to live. Improving of people's living standard and increasing rapidly year by year of the per capita car ownership bring heavy burden to the urban road traffic. In addition, the urban traffic demand is complex and there are many kinds of traffic modes mixing in the city, the road network as a carrier of carrying these traffic modes is facing with enormous pressure. However, because the existing road network in the city limited by geographical space constraints, in order to meet the normal traffic demand, it must give full play effectiveness of existing roads and ensure that these roads are complete and effective. Therefore, it is particularly important to make scientific analysis, accurate judgment, reasonable prediction and timely improvement for the current road network. At the same time, in order to realize the comprehensive goal of urban traffic “more faster, better and cheaper”, it must ensure that the road network of the city can provide basic and reliable travel, as well as still guarantee the traveler successful completion of travel task when there is an accident. So in the process of urban development, it can not be underestimated to strengthen research on the reliability of urban road network for purpose of maximizing the stability of urban traffic system and improving the level of traffic service.In the paper, it uses two kinds of topology construction method applied widely of complex network, namely the main method and the dual method, to establish two network models of part road network of Lanzhou, Chengguan District. Using Pajek software compares the different statistical characteristic of the two network models showing, it proves that the road network constructed by the main method is stochastic dynamic model, and that the degree distribution of the road network constructed by the dual method shows power law characteristic and belongs to scale-free network model. It selects the dual method to construct the road network in this paper, and comprehensively analyzes the reliability according to the invulnerability and restorative of road network, by selecting network efficiency and its relative variation as the measurement index. In invulnerability analysis, it respectively uses random attack strategy and selective attack strategy to simulate the road network under attack, in restorative analysis, it also respectively uses selective recovery strategy and random recovery strategy to simulate the road network under recovering after two attack modes, the results show that the dual road network of Lanzhou, Chengguan District puts up a strong adaptability to the random attack and a relative vulnerability to the selective attack, which further verify the dual road network in the paper showing scale free property. Finally, it introduces topological potential theory in the data field, combines the invulnerability and restorative of road network, calculates the topology potential value of reliability of each road in turn to evaluate its relative reliability and looks out the critical sections of the road network.Different with the existing literature, it designs selective attack algorithm not according to the size of the node degree to successive attack in the network, but according to the size of the topology potential value of node degree, such algorithm considers impact of each node with the surrounding nodes from the near to the distant, so that to judge the importance of each node in the network is more accurately. In addiction, it calculates in turn invulnerability relative value and restorative relative value of each node in the road network and substitutes them into improved topology potential formula, obtains that between the node reliability relative value and the size of the node degree is not completely proportional relationship, the conclusion is when there is big difference between the two node degree values, the greater degree the higher reliability of node in the road network, and when the two node degree values are similar or equal, the higher reliability of all surrounding nodes, the higher reliability of this node.