Adaptive Dynamic Programming with Reinforcement Learning on Optimization of Flight Departure Scheduling

The intricacies of air traffic departure scheduling, especially when numerous flights are delayed, frequently impede the implementation of automated decision-making for scheduling. To surmount this obstacle, a mathematical model is proposed, and a dynamic simulation framework is designed to tackle t...

Full description

Saved in:
Bibliographic Details
Main Authors: Hong Liu, Song Li, Fang Sun, Wei Fan, Wai-Hung Ip, Kai-Leung Yung
Format: Article
Language:English
Published: MDPI AG 2024-09-01
Series:Aerospace
Subjects:
Online Access:https://www.mdpi.com/2226-4310/11/9/754
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850258579771097088
author Hong Liu
Song Li
Fang Sun
Wei Fan
Wai-Hung Ip
Kai-Leung Yung
author_facet Hong Liu
Song Li
Fang Sun
Wei Fan
Wai-Hung Ip
Kai-Leung Yung
author_sort Hong Liu
collection DOAJ
description The intricacies of air traffic departure scheduling, especially when numerous flights are delayed, frequently impede the implementation of automated decision-making for scheduling. To surmount this obstacle, a mathematical model is proposed, and a dynamic simulation framework is designed to tackle the scheduling dilemma. An optimization control strategy is based on adaptive dynamic programming (ADP), focusing on minimizing the cumulative delay time for a cohort of delayed aircraft amidst congestion. This technique harnesses an approximation of the dynamic programming value function, augmented by reinforcement learning to enhance the approximation and alleviate the computational complexity as the number of flights increases. Comparative analyses with alternative approaches, including the branch and bound algorithm for static conditions and the first-come, first-served (FCFS) algorithm for routine scenarios, are conducted. Moreover, perturbation simulations of ADP parameters validate the method’s robustness and efficacy. ADP, when integrated with reinforcement learning, demonstrates time efficiency and reliability, positioning it as a viable solution for decision-making in departure management systems.
format Article
id doaj-art-811e95e2fbb64995976cfaeca5b0ffec
institution OA Journals
issn 2226-4310
language English
publishDate 2024-09-01
publisher MDPI AG
record_format Article
series Aerospace
spelling doaj-art-811e95e2fbb64995976cfaeca5b0ffec2025-08-20T01:56:06ZengMDPI AGAerospace2226-43102024-09-0111975410.3390/aerospace11090754Adaptive Dynamic Programming with Reinforcement Learning on Optimization of Flight Departure SchedulingHong Liu0Song Li1Fang Sun2Wei Fan3Wai-Hung Ip4Kai-Leung Yung5Department of Air Traffic Management, Civil Aviation University of China, Tianjin 300300, ChinaState Key Laboratory of Air Traffic Management System, Nanjing 210014, ChinaDepartment of Science, Civil Aviation University of China, Tianjin 300300, ChinaDepartment of Computer Science and Technology, Civil Aviation University of China, Tianjin 300300, ChinaDepartment of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong SAR, ChinaDepartment of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong SAR, ChinaThe intricacies of air traffic departure scheduling, especially when numerous flights are delayed, frequently impede the implementation of automated decision-making for scheduling. To surmount this obstacle, a mathematical model is proposed, and a dynamic simulation framework is designed to tackle the scheduling dilemma. An optimization control strategy is based on adaptive dynamic programming (ADP), focusing on minimizing the cumulative delay time for a cohort of delayed aircraft amidst congestion. This technique harnesses an approximation of the dynamic programming value function, augmented by reinforcement learning to enhance the approximation and alleviate the computational complexity as the number of flights increases. Comparative analyses with alternative approaches, including the branch and bound algorithm for static conditions and the first-come, first-served (FCFS) algorithm for routine scenarios, are conducted. Moreover, perturbation simulations of ADP parameters validate the method’s robustness and efficacy. ADP, when integrated with reinforcement learning, demonstrates time efficiency and reliability, positioning it as a viable solution for decision-making in departure management systems.https://www.mdpi.com/2226-4310/11/9/754adaptive dynamic programmingdeparture schedulingconstrained integer optimizationsystem robustness
spellingShingle Hong Liu
Song Li
Fang Sun
Wei Fan
Wai-Hung Ip
Kai-Leung Yung
Adaptive Dynamic Programming with Reinforcement Learning on Optimization of Flight Departure Scheduling
Aerospace
adaptive dynamic programming
departure scheduling
constrained integer optimization
system robustness
title Adaptive Dynamic Programming with Reinforcement Learning on Optimization of Flight Departure Scheduling
title_full Adaptive Dynamic Programming with Reinforcement Learning on Optimization of Flight Departure Scheduling
title_fullStr Adaptive Dynamic Programming with Reinforcement Learning on Optimization of Flight Departure Scheduling
title_full_unstemmed Adaptive Dynamic Programming with Reinforcement Learning on Optimization of Flight Departure Scheduling
title_short Adaptive Dynamic Programming with Reinforcement Learning on Optimization of Flight Departure Scheduling
title_sort adaptive dynamic programming with reinforcement learning on optimization of flight departure scheduling
topic adaptive dynamic programming
departure scheduling
constrained integer optimization
system robustness
url https://www.mdpi.com/2226-4310/11/9/754
work_keys_str_mv AT hongliu adaptivedynamicprogrammingwithreinforcementlearningonoptimizationofflightdeparturescheduling
AT songli adaptivedynamicprogrammingwithreinforcementlearningonoptimizationofflightdeparturescheduling
AT fangsun adaptivedynamicprogrammingwithreinforcementlearningonoptimizationofflightdeparturescheduling
AT weifan adaptivedynamicprogrammingwithreinforcementlearningonoptimizationofflightdeparturescheduling
AT waihungip adaptivedynamicprogrammingwithreinforcementlearningonoptimizationofflightdeparturescheduling
AT kaileungyung adaptivedynamicprogrammingwithreinforcementlearningonoptimizationofflightdeparturescheduling