Open AccessCustomized bus scheme design of large transport terminals with jointly optimization of departure time, vehicle allocation and routing
Author(s) -
Wu Yuelin,
Yuan Zhenzhou,
Xiao Qingyu,
Yang Dong
Publication year - 2023
Publication title -
iet intelligent transport systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.579
H-Index - 45
eISSN - 1751-9578
pISSN - 1751-956X
DOI - 10.1049/itr2.12240
Subject(s) - schedule , vehicle routing problem , scheme (mathematics) , routing (electronic design automation) , computer science , service quality , cluster analysis , grid , service (business) , service level , engineering , transport engineering , mathematical optimization , computer network , mathematics , mathematical analysis , statistics , economy , geometry , machine learning , economics , operating system
The customized bus (CB) service of large transport terminals can provide passengers with convenient transfers and door‐to‐door services, which has the potential to help ease the pressure of arriving passenger flow in large transport terminals. Considering the costs of operators and passengers, this paper establishes a multi‐commodity network flow optimization model to simultaneously obtain departure schedule, passenger‐to‐vehicle assignment and routing under the condition of multiple vehicle types. It provides more passengers with high‐quality CB service and obtains higher vehicle capacity utilization. Accordingly, the solution is proposed that combines a branch‐and‐cut algorithm with grid‐density‐based clustering. An illustrative example verifies the effectiveness and tests the impact of the departure time‐window. The case study of Chengdu city compares the results under different types of vehicle schemes and conducts the sensitivity analysis of penalty and minimum load ratio. The experiments conclude: (i) The proposed algorithm stably improves computation speed (76.38%) without affecting the optimal results; (ii) a loose time‐window can serve more people, but the excessively loose one no longer has impacts on the results. (iii) The multi‐type of vehicle scheme performs best compared to the single type of vehicle schemes.