The Application Of Complex Networks In Urban Bus Transport Systems

With the development of urban traffic, urban public traffic systems have already been given more and more attention. In this paper, we create the models of urban public traffic networks, on the basic theory of complex networks, in which obtain the data of bus routes and stations of Shijiazhuang and Nanjing. In order to find advantages and disadvantages of urban public traffic systems, we analyze and compare the statistical properities of three public traffic networks in two different cities.This first chapter briefly describes the research history, concepts and models of complex networks, and the motivation of this study described. Then, in the second chapter describes the situation of traffic networks, the models of traffic networks and research methods. On this basis, then in the third chapter the two models (Space P and Space L) of Nanjing and Shijiazhuang public traffic networks topology features: degree distribution, clustering coefficient, shortest path distribution and betweenness distribution were analyzed and compared in two cities, and a preliminary discussion of the degree-dependent local clustering, a measures of centrality based on betweenness. After statistics we find that in the Space P degree distribution follows an exponential distribution, with the properties of small-world networks; clustering spectrum follows a power-law distribution, with the properties of the scale free networks; while in the Space L the cumulative degree distributions obeys a power-law distribution, with the properties of scale-free networks; betweenness is proportional to the number of node degree and the same as degree in the identification of central nodes. Second, considering the complexity of the real characteristics of public traffic networks, in the fourth chapter we establish a weighted network model (Space R) based on the Space L and Space P. We show the distribution of weight and vertex strength, clustering coefficient in Space R. We explore that the distribution of weight and vertex strength distribution follows the exponential distribution. At the end of fourth chapter we discuss the structure of two cities: the relationship of weighted clustering coefficient and node degree. By calculating between the strength and degree, the strength is also an appropriate standard of centrality. In this paper, we use the weighted clustering coefficient defined by Barrat. In the discussion of the correlation between weighted clustering coefficient and degree, the two cities in the structural organizational of the weighted networks are materially different: approximate a power-law distribution in Nanjing, while Shijiazhuang appears haphazard. Analyzing the reasons and comparing to the structure of Nanjing: Shijiazhuang bus lines are distributed unevenly, urban centers are over-centralized, suburban bus lines are lack, and the degree distribution of transit hub is too large. There is a hidden danger of paralysis and traffic congestion. Finally, through the third and fourth chapters we discuss and analyze the same and difference of the network structure between two cities and the reasons of generate difference, in the fifth chapter we derive the final conclusion.