LAUSR

Abstract Congestion in the air traffic network is a problem with an increasing relevance for airlines costs as well as for airspace safety. One of the major issue is the limited capacity of the air network, that is the limited maximum number of aircraft that can coexist on some points of the network. This value can also vary during the day for safety constraints (weather and visibility). In this work an Autonomous Agent approach is proposed to solve in real time the problem of air traffic congestion. The air traffic network is divided in different sectors each one modeled with a graph. Each sector has its own control decision agent managing the air traffic involved in it. When a congestion occurs in a sector the control decision agent has to reschedule and/or reroute the flights with the aim of respecting capacity constraints and of limiting delays. In this paper rerouting solutions are obtained by applying an iterative procedure based on a generalized Dijkstra algorithm on the graph. We test the effects of different levels of cooperation between agents controlling adjacent sectors while solving the congestion problem.