The modernization of the National Air Traffic Control System is on the horizon, and with it, the possible introduction of autonomous air vehicles into the national airspace. Per the FAA Aerospace Forecast (FAA, 2013), U.S. carrier passenger traffic is expected to average 2.2 percent growth per year over the next 20 years with government statistics indicating that the average domestic load factor for airlines in 2014 was approximately 84.4 percent (US Department of Transportation, 2015). Adding to that demand, the potential introduction of unmanned and autonomous air vehicles motivates reconsideration of control schemes. One of the proposed solutions (Eby, 1994) would involve a decentralized control protocol. Equipping each aircraft with the information necessary to navigate safely through integrated airspace becomes an information sharing problem: how much information about other aircraft is required for a pilot to safely fly the gamut of a heavily populated airspace and what paradigm shifts may be necessary to safely and efficiently utilize available airspace? This thesis describes the development of a tool for testing alternative traffic management systems, centralized or decentralized, in the presence of uncertainty. Applying a computational fluid dynamics-inspired approach to the problem creates a simulation tool to model both the movement of traffic within the airspace and also allows study of the effects of interactions between vehicles. By incorporating a Smoothed Particle Hydrodynamics (SPH) based model, discrete particle aircraft each carry a set of unique deterministic and stochastic properties. With this model, aircraft interaction can be studied to better understand how variations in the nondeterministic properties of the system affect its overall efficiency and safety. The tool is structured to be sufficiently flexible as to allow incorporation of different collision detection and avoidance rules for aircraft traffic management. / Master of Science

Afin de pouvoir subvenir aux futurs besoins en matière de transport aérien il est nécessaire d'augmenter la capacité de l'espace aérien. Les contrôleurs aériens, qui occupent une place centrale dans la gestion du trafic, doivent quotidiennement faire face à des situations conflictuelles (conflits) lors desquelles deux vols risquent de violer les normes de séparation en vigueur si aucune modification de trajectoire n'est envisagée. La détection et la résolution des conflits potentiels contribuent à augmenter la charge de travail des contrôleurs et peuvent potentiellement les conduire à diriger les vols vers des zones moins denses de l'espace aérien, induisant a posteriori un retard pour les vols. Le problème de la capacité de l'espace aérien peut donc être abordé en régulant les flux de trafic de façon réduire la quantité de conflits aériens. L'objectif de cette thèse est de mettre au point une méthodologie destinée à minimiser les risques de conflits aériens en modifiant légèrement les vitesses des appareils. Cette approche est principalement motivée par les conclusions du projet ERASMUS portant sur la régulation de vitesse subliminale. Ce type de régulation a été conçu de façon à ne pas perturber les contrôleurs aériens dans leur tâche. En utilisant de faibles modulations de vitesse, imperceptibles par les contrôleurs aériens, les trajectoires des vols peuvent être modifiées pour minimiser la quantité totale de conflits et ainsi faciliter l'écoulement du trafic dans le réseau aérien. La méthode retenue pour mettre en œuvre ce type de régulation est l'optimisation sous contrainte. Dans cette thèse, nous développons un modèle d'optimisation déterministe pour traiter les conflits à deux avions. Ce modèle est par la suite adapté à la résolution de grandes instances de trafic en formulant le modèle comme un Programme Linéaire en Nombres Entiers. Pour reproduire des conditions de trafic réalistes, nous introduisons une perturbation sur la vitesse des vols, destinée à représenter l'impact de l'incertitude en prévision de trajectoire dans la gestion du trafic aérien. Pour valider notre approche, nous utilisons un outil de simulation capable de rejouer des journées entières de trafic au dessus de l'espace aérien européen. Les principaux résultats de ce travail démontrent les performances du modèle de détection et de résolution de conflits et soulignent la robustesse de la formulation face à l'incertitude en prévision de trajectoire. Enfin, l'impact de notre approche est évalué à travers divers indicateurs propres à la gestion du trafic aérien et valide la méthodologie développée. / As global air traffic volume is continuously increasing, it has become a priority to improve air traffic control in order to deal with future air traffic demand. One of the current challenges regarding air traffic management is the airspace capacity problem, which is acknowledged to be correlated to air traffic controllers' workload. Air traffic controllers stand at the core of the traffic monitoring system and one of their main objective is to ensure the separation of aircraft by anticipating potential conflicts. Conflict detection and resolution are likely to increase workload and may lead them to reroute aircrafts to less dense areas, triggering off flight delay. The airspace capacity problem can hence be tackled by regulating air traffic flow in order to reduce the global conflict quantity. The objective of this thesis is to develop a methodology aiming at minimizing potential conflicts quantity by slightly adjusting aircraft speeds in real time. This approach is mainly motivated by conclusions of the ERASMUS project on subliminal speed control, which was designed to keep air traffic controllers unaware of the ongoing regulation process. By focusing on low magnitude speed modulations, aircraft trajectories can be modified to reduce the quantity of conflicts and smoothen air traffic flow in the airspace network. The method used to carry out this type of regulation is constraint optimization. In this thesis, we develop a deterministic optimization model for two-aircraft conflicts which is then adapted to large scale instances using Mixed-Integer Linear Programming. In order to reproduce realistic navigation conditions, uncertainty on aircraft speeds is introduced with the goal of modeling the impact of trajectory prediction uncertainty in air traffic management. To validate our approach, a simulation device capable of simulating real air traffic data over the European airspace is used. Main results of this work reveal a significant conflict quantity reduction and demonstrate the robustness of the developed model to the uncertainty in trajectory prediction. Finally, the impact of our model on air traffic flow is measured through several air traffic management indicators and validates the proposed methodology.

No description available.

The problem of aircraft conflict detection and resolution (CDR) in uncertainty is addressed in this thesis. The main goal in CDR is to provide safety for the aircraft while minimizing their fuel consumption and flight delays. In reality, a high degree of uncertainty can exist in certain aircraft-aircraft encounters especially in cases where aircraft do not have the capabilities to communicate with each other. Through the use of a probabilistic approach and a multiple model (MM) trajectory information processing framework, this uncertainty can be effectively handled. For conflict detection, a randomized Monte Carlo (MC) algorithm is used to accurately detect conflicts, and, if a conflict is detected, a conflict resolution algorithm is run that utilizes a sequential list Viterbi algorithm. This thesis presents the MM CDR method and a comprehensive MC simulation and performance evaluation study that demonstrates its capabilities and efficiency.

No description available.