Integrated optimization of train timetable and train unit circulation for a Y-type urban rail transit system with flexible train composition mode

With the increasing complexity of urban rail transit (URT) networks and the revolution of train control technology, a large number of diverse operation and organization modes have emerged. This paper studies the train operation plan problem for a Y-type URT system with flexible train composition mode. The studied problem can be defined as an integrated optimization of the train timetable problem (TTP) and train unit circulation problem (TUCP) under time-dependent passengers, where the train compositions can be coupled or decoupled during the trip. By considering train operation constraints, train composition constraints, train unit circulation constraints, and passenger demand constraints with limited train capacity, our problem is formulated as an integer programming (IP) model, where the objective is to minimize the passengers’ total waiting time and the number of train units used. To solve large-scale real-world instances, we design a heuristic algorithm by improving the classical framework of the adaptive large neighborhood search (ALNS) algorithm. Several groups of experiments, including small-case instances and real-world instances of Guangzhou Metro Line 14, are investigated to verify the effectiveness of the proposed model and algorithm. Compared with the traditional fixed train composition mode, both the passengers’ total waiting time and the number of train units used under the condition of flexible train composition mode are reduced. This will effectively improve the quality of the passenger service and reduce the costs of URT departments.