.png){kind=icon}
返回
Air traffic controller scheduling
.png){kind=icon}
PRE
AI
DOI：10.1016/j.cie.2024.110123.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
This paper investigates the air traffic controller scheduling problem in which each employee must be assigned up to 10 shifts over a two-week planning horizon. Additional assignments include a 30-minute lunch break and several 15-minute rest breaks within each shift. The goal is to select the minimum number of employees that balances a weighted combination of demand undercoverage, overcoverage and labor costs. Additional constraints include the minimum/maximum number of employees, minimum shift and demand coverage, minimum number of consecutive working hours, bounds on working hours, minimum days off, maximum consecutive days on, and minimum rest time between shifts. A 2-phase algorithm is developed to efficiently construct weekly schedules for the controllers. In Phase 1, employee-shift assignments are made with a randomization algorithm and a greedy post-processor. In Phase 2, 15-minute breaks are added to the selected shifts by solving a second optimization model. The proposed methodology is tested on both small instances defined by the Federal Aviation Administration in their 2022 challenge, and large instances with up to 52 employees. Compared to the actual schedule, the results verify that the 2-phase algorithm offers significantly better schedules with respect to all metrics. We conclude with a parametric analysis of the number of employees to determine how staffing levels effect solution quality.
Keyword:
Air traffic controller scheduling
Mixed-integer linear programming
Greedy algorithm
Randomization algorithm
Lunch breaks
Rest breaks
Parametric analysis
期刊
IF:
6.5
论文数:
1.0W
被引数:
3.8W
机构
引用论文
.png){kind=icon}
查看更多内容