The Study On Metro Path Choice Model And Its Application Based On Ant Colony Algorithm

As the development of China's urban metro network is getting faster day by day, the characteristics of passenger flow under the network environment also appear. Moreover, some passengers have several paths to choose when they ride metro. Therefore, the spatial and temporal distribution characteristics and the behaviors of path choice in metro network have changed. Generally, passengers will consider in-vehicle time, transfer time, transfer number and fare when choosing their travel paths. Except for these four factors, the congestion level of the metro line was introduced into this research in this thesis. Thus, the optimal travel path can be chosen from a practical view, which can guide passenger flow, improve the comfort in the metro system for passengers, and make good use of metro system and enhance its service level.Firstly, this thesis reviewed the literature on path choice algorithms and systems. Pros and cons of existing studies were summarized.Secondly, Dijkstra algorithm, Genetic algorithm and Ant colony algorithm, which are common-used in nowadays, were introduced. Pros and cons of each algorithm were analyzed. Ant colony algorithm is chosen as the shortest path algorithm, which will be developed in this thesis considering the demand of metro network. Thirdly, in-vehicle time, transfer time, transfer number, fare and congestion, which affect metro path choice, was investigated, analyzed, and quantified. A metro path choice model and solving algorithm were developed.Fourthly, the metro path choice system based on Ant colony algorithm is developed with VB and GIS tool based on the designed system objective, in which database and core algorithm were built.Finally, a case-study network was chose. The parameters of Ant colony algorithm were discussed and decided in order to improve algorithm efficiency. The developed system was used to search the optimal path and the results were analyzed in this case-study network.