Dynamic Passenger Demands Oriented Optimization of Virtually Coupled Train Formation Scheduling

Haobei Zhang,Tao Tang,Hongjie Liu,Yinghui Lang
DOI: https://doi.org/10.1109/cac57257.2022.10056085
2022-01-01
Abstract:The research on virtual coupling technology in metros is a hot topic recently. Intelligent automatic train units can flexibly form virtually coupled train formations (VCTFs) when needed, so that the train capacity would be able to adapt to passenger demands, which are changing dynamically with time and space. However, matching the capacity of trains with passenger demands is a difficult decision-making optimization problem. In this work, we propose a collaborative control problem to decide when and how to create a VCTF to meet the dynamic passenger demands in a metro system. A mathematical model is formulated which can simultaneously optimize the number of related train units in a VCTF, the train dwelling time at a station and the running time in a section. Performance indicators such as train load factor and number of stranded passengers are adopted to measure the satisfaction of the operators and passengers. Then, a multi-objective genetic algorithm is adopted to obtain the Pareto optimum solutions. Finally, simulation experiments are conducted, and the relationship between line efficiency and passenger demand in different discrete time periods is analyzed to verify the correctness and feasibility of the model. Compared with the traditional timetabling method, adopting the virtual coupling technology can better adapt the train capacity to meet the time changing passenger demands, which is welcomed by both metro operators and passengers in reality.
What problem does this paper attempt to address?