Section: Application Domains

Air Traffic Control

Air Traffic Control has been an application field of Marc Schoenauer’s work in the late 90s (PhD theses of F. Médioni in 1998 and S. Oussedik in 2000). It was revived recently with Gaëtan Marceau-Caron’s CIFRE PhD together with Thalès Air Systems (Areski Hadjaz) and Thalès TRT (Pierre Savéant), tackling the global optimization of the traffic in order to increase the capacity of the airspace without overloading the controllers. A new formulation of the problem, modeling the plane flows with Bayesian Networks, has been proposed to the Air Traffic Control community [48] , [50] . The goal of the optimization is to minimize the cumulated delays of all flights, while maintaining a reasonnable level of congestion in all sectors. These objectives are computed using Monte-Carlo simulations of the Bayesian network, and Evolutionary Algorithms are used to address the resulting stochastic multi-objective optimization problem [49] .