Research On The Optimization Of Traffic Organization For Urban Rail Transit Network

With the rapid development of urban rail transit in China, the networking of urban rail transit becomes an inevitable trend. The networking process is building step by step. In different stages in the networking progress, the problems appearing are also different. The planning, design, construction and operation of line network are all in relation to networking. With more and more lines are in operation, it is also more and more important and complex to achieve networking traffic organization. Arranging train schedules appropriately according to the characteristics of passenger flow benefits passengers not only by providing better service and more comfortable experience but also reducing commute time. It is also of benefit to enterprises by reducing the cost and the pressure of passenger flow and improving the operation efficiency. It is very important for the sustainable development of urban rail transit in China. Therefore, this paper studied the following aspects, put forward the corresponding solutions, built optimization model, designed algorithms and simulated using MATLAB, details were as follows:1. The concept of urban rail transit was elaborated overall and the connotation and significance of networking was analyzed. It also discussed the optimization theory of networking traffic organization and studied the optimization algorithm for solving integer programming problems, including genetic algorithm and multi-objective evolutionary algorithm.2. The diversification in train route operation organization was analyzed, especially the organization scheme of long-short routing operations was analyzed in depth according to the complex demands of passengers of the networking conditions. Through establishing the optimal model of train operation scheme, the parameters of long-short routing were determined, including running interval and cycle time of long-short routing, selection or re-entry point position of short routing, re-entry time at the intermediate station and the number of rolling stock. The model took running interaction between long and short routings into condition on the premise of safety and the demand of passenger flow. The long routing running interval was used to adjust the short routing running interval and driving cycle time, and the restrictive relationships between parameters were described. In this paper, taking Chengdu Metro Line 2 as example, the model was used for calculation, and the results showed that the model could achieve optimization purposes.3. The coordination and optimization problems for networking timetable in transfer station were studied during the flat peak period. Combining the traffic organization and coordination in the transfer station with the traffic organization and coordination between the transfer stations, the connecting optimization model for trains’arrival timetable and leaving timetable in a single transfer station and the coordination and optimization model for networking timetable were established respectively. During the coordination progress, the weight of each transfer station in the network was distributed based on graph theory, making sure of the coordination and optimization results more effective. Genetic algorithm and multi-objective and non-dominated sorting evolutionary algorithm were designed based on superior insurance strategy, and the optimization models in a single transfer station and networking stations were solved respectively. Through verification, the models and algorithms could effectively reduce the waiting time of passengers in transfer stations.4. The coordination and optimization for networking timetable in transfer station were studied during the rush hours. Applying the idea of train connecting coordination during flat peak period to rush hours, and analyzing the factors that had influence on passenger flow in each platform in the transfer station, formulas to describe the variation of waiting passengers with time were obtained. The optimization model which minimized the maximum of total waiting passengers during the study period was established. In the modeling process, the case was considered where passengers can’t board on the first arrival train that they encountered. Genetic algorithm based on superior insurance strategy was designed. Through verification, it could reduce the maximum waiting passengers effectively by appropriate train connecting coordination in the transfer station.5. The optimization schemes for passenger flow in transfer station during rush hours were analyzed comprehensively. Aiming at the coordination scheme for transport capacity, and based on the characters that trains run cyclically, recursion relations were established and then the relational model was obtained to calculate the number of waiting passengers, the number of passengers taken away and the remaining passengers in transfer sites when trains arrived and left in transfer station. Based on the model referred above, we used the optimized method to coordinate the transportation capacity, and solved it by Monte Carlo simulation. By designing the basic parameters and case analysis, the coordination schemes of traffic intervals and train capacities were obtained.