Abstract: In this article an improved Flow Management tool is proposed and evaluated. In most research currently performed on the subject of air traffic flow management, Mixed Integer Linear Programming (MILP) techniques are used to come to a solution. In the approach proposed herein, a Petri-net algorithm is used to model the air traffic flow. By iteratively looking ahead in time, future reservations on capacity are made. In-air waiting times are then suppressed by imposing pre-departure delays, based on the maximum delay that a flight can expect at a node along its route. The newly developed algorithms were shown to effectively suppress in-air waiting time by imposing pre-departure delays. Second-order (knock-on) network effects are also included, leading to a more efficient utilisation of capacity. Moreover, in a large European-wide scenario with disruption at several major hub airports, prioritisation of flights to and from these airports was shown to considerably alleviate the delays at the impacted airports. While this did come at a small cost to the total duration of imposed delays, total monetary cost of delay was found to be reduced.
Keywords: Air traffic management (ATM), Air Traffic Flow Management (ATFM), flow and capacity management, network analysis, optimisation, prioritisation, petri-net analysis, model-based simulations
