Train platooning, which allows multiple train units to be virtually coupled into a platoon with very short following distances, has become an emerging technology in railway industry. Our study investigates the energy-efficient operation of train platoons to reduce the energy consumption of trains while satisfying time-varying passenger demands in practice. Specifically, we construct a mixed integer programming (MIP) model in order to simultaneously optimize the timetable and speed profiles of train platoons in an urban rail transit line. The aim of the developed MIP model is to minimize the total energy consumption of trains as well as the waiting time of traveling passengers, while satisfying the practical constraints, such as the train movement constraints, passenger loading constraints for different train platoons, etc. In order to solve the MIP model more efficiently, we develop a hybrid iterative algorithm that combines a Deep Q-Network (DQN) and an MIP solver, in which the DQN generates energy-efficient speed profiles for different types of platoons and the MIP solver then determines the optimal train schedule. Finally, we conduct real-world case studies on the Beijing subway Yizhuang line to verify the effectiveness of our approach. The results indicate that the energy-efficient operation of train platoons can achieve a better tradeoff between service quality and system energy consumption.
Integrated Optimization of Energy-Efficient Timetable and Speed Profiles for Train Platoons in Urban Rail Transit Systems
2024-09-24
576956 byte
Conference paper
Electronic Resource
English
Integrated energy‐efficient optimization for urban rail transit timetable
Wiley | 2023
|European Patent Office | 2023
|Train timetable adjusting method under sudden interruption scene of urban rail transit
European Patent Office | 2023
|Taylor & Francis Verlag | 2023
|