Study On Urban Transit Network Generation And Optimization Based On Complex Network Theory

China’s automobile industry had obtained an unprecedented developmentduring the global economic crisis. The urban trips structure had undergone profoundchanges. The proportion of car trips had obtained a significant increase. Residents′travel mobility had improved. These had caused roads blocked and the phenomenonis spreading from initial Beijing, Shanghai, Guangzhou to second-tier cities. NowTraffic congestion has been common things in the cities and it has been more andmore serious. The scholars have to solve the problem that how to relieve and governurban traffic congestion. Vigorously developing urban public transport to ease urbantraffic congestion has become a consensus. But urban public transport, especially thecommon bus, has to share the road resources with cars. Therefore traffic congestionhas comparatively large effects on it. In order to make urban public transporteffectively deal with traffic congestion and play to its maximized advantages wehave to reasonably generate and optimize the urban public transport network.Urban public transport network generation and optimization are complexstudies. Multi scholars used complex network theory for the complexity of the urbanpublic transport network after the beginning of its studies. They explored itsevolution mechanism and formation mechanism. But for transit network generationand optimization these scholars studied the distribution law of its complex statisticalindicators, qualitatively analyzed its problems and proposed measures to optimizethe network. They rarely study the bus network topology characteristics and thenetwork mathematical models. When generating and optimizing bus network,consider the effects of traffic congestion on the public transport system and residentstravel demand. The existing literature has not yet targeted research on it.Therefore the paper relies on the National Natural Science Foundation of“Research on the key models and its algorithms in urban transit network integrateddesign（51078168）” to study transit network on the base of domestic and foreignstudies. Make the station research as an entry point, Three models were established,namely complex public transport network model based on efficiency-driven, complex public transport network model based on gravitation-driven and urbantransit network optimization model based on the edge betweenness. Then urban busstop spacing optimization was studied. The completed research results are asfollows:1. Make urban bus station as research center. Firstly analyze urban bus station,then analyze station demand characteristics on the base of Fractal Theory. Theresults show that when the statistics interval is2min the station demand haveapparent fractal characteristics. Analyze bus stop type, then according to the trafficdetermine the number of first and last stop, the middle stop and their treatmentmethods. Apply double constrained gravity model to predict the bus stops OD matrixwith obtaining traffic zone OD matrix.2. Study the basic issue of urban public transport network generation andoptimization. Design k shortest path algorithm with k-1shortest path on the base ofDijkstra algorithm. Analyze planning objectives and constraints of urban publictransit network planning, build urban public transport efficiency networkinfrastructure model based on station capacity restrictions.3. In order to obtain best urban transit network, analyze three typical complexnetworks of random networks, small world networks and scale-free networks. Whenlarge passenger demand appear in the network the scale-free networks get theadvantage because of their topological properties. Analyze classic BA model fromBarabasi and Albert, propose two complex public transport network generationalgorithms based on growth mechanisms and preferences mechanism, namelycomplex public transport network generation algorithm based on efficiency-drivenand complex public transport network generation algorithm based on thegravity–driven. They all follow the maximized transport efficiency and networkcapacity as bus networks generation targets. The difference is that the former wasdriven by transport efficiency and the latter had considered the impedance betweenthe stations and the line direction, it is driven by the gravitation between the stations.4. In order to relieve traffic congestion on the public transport system in theurban public transport network optimization this research introduced relevantparameters to simulate the residents travel strategy changes in the traffic congestionby analyzing the residents travel strategy and expanding the edge betweenness, Thendefine the effective weight, the effective path, the effective lines and the effectivenetwork to build the urban public transport network optimization model and its algorithm based on edge betweenness. It also determined the physical meaning of themodel parameters and their calibration by analyzing the relationship between themodel parameters and network indicators such as zero-flow network efficiency, thecrowded network efficiency, network efficiency loss, the average travel time, theaverage trip distance, the average trip speed etc.5. The paper proposed a two-step method which could calculate the actualstation spacing to combine the advantage of mathematical models and experience toset up stations. The first step is to establish the station spacing optimal model tocalculate the optimal station spacing under the ideal conditions based on modelassumes. The second step is to analyze the affect factors of station spacing andclassify them into five combination factors, namely the equivalent acceleration, thespace mean speed of the vehicle, the parking time of vehicle at stop, averagepassenger trip distance and average arrival speed of passenger. This researchcompleted the setting of urban bus stops through building a station spacingcorrection model to correct the optimal station spacing.6. By learning from the related outcomes about the project of “Public transportplanning in Changchun Economic and Technological Development Zone(2011-2020)”, the paper made the empirical study on the proposed model andmethod. Firstly analyze the current situation of public transit in Changchun,determine the number and layout of the first and last station and middle station, thenpredict station requirements. Next generate new lines in Changchun and optimizetheir existing unreasonable lines. Finally re-optimized the station spacing, evaluatethe generation and optimization results of transit network in Changchun Economicand Technological Development Zone.The results of this research can expand the range of applications of complexnetwork theory, rich urban public transport planning theory system, improve theresponse capacity of urban public transport network to traffic congestion, promotethe healthy and orderly development of the urban public transport system, have hightheoretical value and practical significance.