The urban rail transit line operating in the express and local train mode can solve the problem of disequilibrium passenger flow and space and meet the rapid arrival demand of long-distance passengers. In this paper, the Logit model is used to analyze the behavior of passengers choosing trains by considering the sensitivity of travel time and travel distance. Then, based on the composition of passenger travel time, an integer programming model for train stop scheme, aimed at minimizing the total passenger travel time, is proposed. Finally, combined with a certain regional rail line in Shenzhen, the plan is solved by genetic algorithm and evaluated through the time benefit, carrying capacity, and energy consumption efficiency. The simulation result shows that although the capacity is reduced by 6 trains, the optimized travel time per person is 10.34 min, and the energy consumption is saved by about 16%, which proves that the proposed model is efficient and feasible.
1 December 2016 in Journal of Advanced Transportation
The metro passenger route choice, influenced by both train schedule and time constraints, is important to metro operation and management. Smart card data (Automatic Fare Collection (AFC) data in metro system) including inbound and outbound swiping time are useful for analysis of the characteristics of passengers’ route choices in metro while they could not reflect the property of train schedule directly. Train schedule is used in this paper to trim smart card data through removing inbound and outbound walking time to/from platforms and waiting time. Thus, passengers’ pure travel time in accord with trains’ arrival and departure can be obtained. Synchronous clustering (SynC) algorithm is then applied to analyze these processed data to calculate passenger route choice probability. Finally, a case study was conducted to illustrate the effectiveness of the proposed algorithm. Results showed the proposed algorithm works well to analyze metro passenger route choice. It was shown that passenger route choice during both peak period and flat period could be clustered automatically, and noise data are isolated. The probability of route choice calculated through SynC algorithm can be used to revise traditional model results.
Metro trains stop operation during midnight for the maintenance of vehicles and tracks in most cities, where passengers heavily rely on the metro for their daily life. Thus, passengers may miss the last trains when they travel by metro at late night. However, the last trains are especially important because they are the last chances for many passengers to travel by metro. If passengers miss the last trains, they have to choose buses, taxies, or other transport modes to complete their trips. Consequently, it is necessary to optimize the schedule of the last train coordination to meet the demand of passengers at transfer stations during midnight. The passenger demand for last trains is a vital input to deal with the coordination of last train transfers. This paper focuses on forecasting the potential passenger demand of last trains from public transport data including taxi (FCD data) and bus (GPS/smart data) systems. A solution for taxi and bus data is developed to calculate the potential passenger demand for all the transfer directions of the target stations. Then, a model for the coordination of last train transfers based on the potential passenger demand is proposed. The genetic algorithm is applied to solve the model. The effectiveness of the proposed method is evaluated using the Shenzhen metro network with several data on a typical Friday. The research is to provide theoretical guidance and technical reference for the metro operation when to compile the last train schedule. It is supposed to improve the modern operation management level of metro systems.
Passenger travel paths in metro networks have become more diversified with the development of network structures and the complexity of train schedules. Nowadays, passengers may have more than one alternative path in an OD (Origin-Destination) pair. In order to provide high-quality service to passengers, this paper proposes a smart path finding method to recommend fast and comfortable routes to passengers. By including the structure of the metro network as a two-dimensional plane and time as the third dimension, the space-time range of passenger activities is constructed. The accessible transfer stations for arrival are firstly computing by forward searching from O station, and the accessible transfer stations for departure are then identified by backward searching from D station. The intersection of the two feasible trajectories is regarded as the feasible path set of passengers. The path is recommended to passengers using generalized path impedance with the consideration of the situation that passengers have different perceptions of travel time, transfer penalty, and congestion tolerance. Finally, the proposed method is validated with the data of Shenzhen Metro under different traffic demand conditions, including the uncongested condition, the condition with congestion, the last train condition, and the condition with train delays, respectively. The time complexity of the proposed algorithm is also compared with that of some conventional algorithms. The results show that the proposed algorithm has lower algorithm complexity and is more suitable for the increasingly large-scale metro network.
26 December 2023 in IET Intelligent Transport Systems
The rapid advancements in information technology and intelligent systems within urban rail transit (URT) systems have highlighted the need for more personalized route recommendations that consider passengers’ travel habits. This study aims to address this issue by investigating passenger travel routes alongside other passengers who share similar travel preferences, utilizing collaborative filtering (CF) techniques. The approach involves analyzing historical card data to assess passenger travel profiles, including actual travel time under crowded conditions. By considering both individual passenger preferences and the preferences of similar passengers, a CF algorithm is employed to offer personalized route recommendations. The Shenzhen metro is used as a case study to illustrate the proposed method. The results demonstrate that the proposed approach surpasses traditional route recommendation methods by providing tailored suggestions that align more closely with passengers’ travel preferences. These findings emphasize the value of incorporating passenger travel preferences into route recommendation models, thereby enhancing the accuracy and effectiveness of personalized route recommendations within URT systems.