Airspace configuration using air traffic complexity metrics
Résumé
Flow regulation is a critical process in air traffic management, ensuring that the incoming traffic does not exceed the ability of air traffic controllers to handle it safely and efficiently. Currently, the european Flow Management Positions (FMP) use flight counts and sector capacities to assess the traffic load and build predicted opening schemes. These schemes, made of predefined airspace configurations, are used to detect potential overloads. Some past research undertaken at the Global Optimization Laboratory led to think that this process was not grounded on solid scientific notions, as concerns the quantification of the controllers workload. Consequently, it is proposed to stop using flight counts and sector capacities to predict this workload, and to use relevant air traffic complexity metrics instead. Another proposal is to explore all the possible combinations of elementary sectors, instead of the small subset of pre-defined configurations currently being used, so as to offer the maximum capacity to the incoming traffic. In previous works, we assessed the relevance of complexity metrics by comparing their relative influence on the sector status prediction (merged, manned1, or split) made by a neural network. Real sector statuses issued from filed configurations were used to train the neural network2. A fairly simple relationship between the relevant metrics and the sector status was found. The main contribution of this paper is to use the relevant metrics and the sector status prediction to build realistic airspace configurations. The computed configurations are compared to the actual configurations archived by the ATC centers, and to the FMP opening schemes.
Domaines
Optimisation et contrôle [math.OC]Origine | Fichiers produits par l'(les) auteur(s) |
---|
Loading...