Study On Evolution Characteristics And Driving Mechanisms Of Bus Dynamic Network Oriented By Vehicle Scheduling

Urban public transportation plays an important part in the city transportation system. Bus priority as a strategy of transportation development has become an inevitable choice to ease traffic congestion and to achieve sustainable development in many cities. Most researches on public transport network and bus scheduling are based on the static topological transformation of public transportation network and cored with the evaluation of bus operation to establish system optimization model, but the model is complex and difficult to solve. The bus dynamic network, in a certain model mechanism, can be abstracted into a dynamic network structure model through analyzing the dynamic location of the vehicles on the network. The network model can reflect the dynamic characteristics of the location of the buses on the network, so we can study the regulation of dynamic changes, using the theories and methods of complex network, and resolve the evolution regulation of buses driving dynamic characteristics under different conditions. This article is based on the fund project of "The bus dynamic network complexity of vehicle scheduling and its empirical study (No.51308176)", we construct a simulation platform to build bus dynamic network based on VISSIM. This article focuses on exploring the evolution characteristics and driving mechanism of the bus dynamic network through simulation and data analysis, the main contents are:Firstly, after analyzing the structure of bus network, we construct a bus operation simulation platform combining VISSIM with MATLAB. The simulation platform can deal with long-time simulation with different timetables and can export real-time location information of the vehicles. The platform can also build bus dynamic network based on real-time location information of the vehicles automatically. The simulation system can provide operation platform and data support for the research on evolution characteristics and driving mechanism in bus dynamic network.Secondly, we propose the active and un-active, active degree, active rate and un-active rate statistical characteristics through analyzing simulation data. Then based on existing bus dynamic network model we adopt statistical methods to studying the variation regulation of statistical characteristics, not only including different types of nodes of the active and un-active, active degree, but also active rate and un-active rate. Through that, we complete the evolution characteristics research of bus dynamic network in time and space.Again, we propose a node-level driving method based on classification analysis in this paper. Then, on the basis of the existing bus dynamic network model we study the variation regulation of statistical characteristics, not only including the active and un-active of nodes with six bus lines and two bus lines in driving lines but also the active rate in different driving lines bus operation intervals. The results provide practical theory for scientific and efficient bus scheduling.In the bus dynamic network evolution process, the different types of nodes statistical properties and the evolution time have a direct relationship. Thus, we should adapt to take the appropriate scheduling strategy in real time when faced with different driving strategy, to facilitate buses real-time dynamic scheduling and improve the bus network of overall transportation.