.png){kind=icon}
Return
Point merge system using decentralized control
.png){kind=icon}
OA
AI
DOI：10.1016/j.cja.2026.104489.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Point Merge (PM) is a systematic method for sequencing the arrival flows of aircraft in a terminal area. In this paper, a finite state machine-based agent system is used to model the PM system. A decentralized control method is proposed which employs message-sending policies to replace the centralized controller. Three different decentralized policies are proposed, which are based on and inspired by the Self-organizing Time Division Multiple Access (STDMA). Under three control policies, all the arriving aircraft need to send messages to each other to achieve the turn sequencing. For the method validation, a dynamic scenario with two sequencing legs is used to compare these three policies, where the centralized control acts as a benchmark policy. Both sequencing leg capacity and operational results of the proposed policies are investigated. It is found that a decentralized policy using contiguous reservation and a network entry slot can achieve a similar capacity to the centralized control.
Keywords:
Pointmerge
Multi-agent systems
Finite state machine
Decentralized control
Air traffic management
Journal
IF:
5.7
Papers:
4.7K
Citations:
1.4W
Organization
Cited Papers
No cited papers available