Search

Road disruptions frequently occur in practice, resulting in a productivity loss of bus services and widespread passenger delays. These negative impacts are significant, especially when multiple routes are influenced at the same time. In order to mitigate these impacts, this paper proposes a multi-route coordination approach that collaboratively adjusts multiple bus routes and optimizes bus timetables to provide effective alternative bus services for passengers. Three adjusting strategies are adopted for bus routes with varying passenger demand: detouring, short-running, and cancellation. To address the multi-route coordination problem, a column-generation-based two-stage framework is developed. Concretely, a column generation technique is utilized in the first stage to iteratively generate candidate adaptive bus routes and passenger paths. Following that, an integrated integer programming model is built in the second stage to simultaneously determine bus timetables and the combinations of those adaptive bus routes. After realizing the low computational efficiency for large-scale problems, this paper designs a customized decomposition algorithm based on set partitioning to solve the presented model and obtain near-optimal solutions efficiently. Finally, the proposed methodology is applied to an illustrative Sioux Falls network and a real-world bus network in Beijing to verify its validity and effectiveness. Three comparative analyses are conducted to discuss the advantages of the three adjusting strategies, to investigate the benefits of coordinating bus timetables, and to explore the applicability of different adjusting strategies, respectively. • A multi-route coordination problem for bus systems in response to road disruptions. • A column-generation-based two-stage framework to address the problem. • A customized decomposition algorithm based on set partitioning for model solving. • The framework and algorithm are applied to a real-world bus network in Beijing. [ABSTRACT FROM AUTHOR]

Traditional customized buses travel on fixed routes, which cannot satisfy passengers' flexibility and convenience requirements. This paper studies a demand-responsive transit (DRT) service that can continuously adjust the path based on passengers' dynamic demand. The path optimization model is established with more realistic constraints to create a bus travel plan within a specified area, and the model not only considers the preferred time windows of passengers but also maximizes the benefits of the system. Based on simulated annealing, a dynamic genetic algorithm is designed to generate the static initial travel path, and the dynamic travel path is continuously updated to satisfy the real-time demand. To evaluate the proposed model and algorithm, a case study in a typical residential community of Beijing, China, is conducted based on transit smart card records. According to the case study results, the convenience, travel time, and economic and environmental benefits of the DRT service are assessed via comparison with traditional buses and private cars. The analysis results demonstrate the feasibility and significance of the method, and it can be used by transit planners to design a superior DRT service. [ABSTRACT FROM AUTHOR]

User profiles, considered as one of the fundamental inputs of recommendation systems and customized services, can be rationally applied in the public transport domain to represent passengers' characteristics and behavioral preferences. A user profile of a bus passenger, termed as a bus passenger profile (BPP), is an assortment of labels containing passengers' travel features. This paper proposes a data-driven framework for developing BPP so as to provide guidance on how to create and estimate user profiles for bus passengers based on smart card data. The proposed method comprises three steps. (i) Data preprocessing aimed at extracting key information and preparing passenger profiling. (ii) A tag system aimed at storing the estimated travel features of passengers. (iii) Knowledge graphs aimed at connecting various BPP with semantic edges for practical application of prior knowledge in downstream tasks. The developed framework is implemented in a case study of the Beijing bus system. Deployment of the developed framework has demonstrated that it can satisfactorily develop BPP, while prior knowledge from the BPP-based knowledge graphs can benefit downstream tasks. [ABSTRACT FROM AUTHOR]

In recent years, customized bus (CB) service has become an innovative model for public transportation in China. Its ability to provide advanced, timely, personalized, and flexible service has made it very popular. However, previous studies have ignored its diversion effect: How can CB service help conventional buses reduce on-board crowding during rush hours? In this paper, we optimize conventional and CB buses as a whole based on an empirical study in Beijing. We propose a new model to compare the cost of single conventional buses (Single Con) and the integration of conventional and customized buses (Con+ CB). Passenger diversions from conventional buses to CB ones are specifically considered. To do this we explicitly measure the on-board crowding cost of conventional buses by load factor and distance factor. Real-world data from Beijing, China, are employed to assess the model. We obtained the departure arrangements and the corresponding costs under different demand cases. The results are analyzed and compared. Sensitivity analyses are then conducted to study the effects of line-haul distance and load factor on cost. The results show that (1) adding new modes of transportation does not always significantly reduce the total cost of bus service; (2) CB buses can effectively ease congestion on conventional buses only at high demand densities but not at low densities; (3) crowding cost increases caused by the distance factor far exceed those caused by the load factor. • Explicitly measure crowding cost of conventional buses. • Propose cost models of Single Con buses and integration of Con+ CB buses. • Employ the real-world data in Beijing to assess the model. [ABSTRACT FROM AUTHOR]