Metro crew planning with heterogeneous duty paths and period-cycle pattern considerations

Feng Xue,Xinyu Zhang,Ping Hu,Xiaochen Ma,Chongshuang Chen

DOI: https://doi.org/10.1016/j.cie.2023.109354

2023-06-10

Abstract:A metro crew plan consists of a crew scheduling plan and crew rostering plan. Based on the shift mode of working in three shifts with four groups in the fixed-shift system of Chinese metro operating enterprises, this paper considers the nature of the work of metro crews to study the equitability of metro crew planning. The train running diagram is split into a sequence of segments by the duty points, including shifting stations, depots and parking lots. The connections between segments are transformed into a time-space-state network based on the crew duty path, which is the path that crew members take while clocking in and out. The connections between shifts and the matching of crew members and shifts are embedded in the duty path chain cycle. The alternating direction method of multipliers and an improved particle swarm optimization algorithm are used to design the corresponding solution algorithm, and the metro crew plan problem can be solved by the proposed method, which was successfully tested on real-life data from the Chengdu Metro. The results of the study show that our models and methods yield 134 crew shifts, including 45 morning shifts, 42 day shifts and 47 night shifts, with an optimization rate of 15.7%. A 9.37% increase in crew duty efficiency with an optimization rate of 13.17% is observed. Additionally, the optimization rates of the total work intensity equitability and the salary equitability of each crew member reach 20.23% and 34.7%, respectively.

computer science, interdisciplinary applications,engineering, industrial