Archive ouverte de l'ENAC
Not a member yet
3458 research outputs found
Sort by
Open wind tunnel experiments of the DarkO tail-sitter longitudinal stabilization with constant wind
No full textInternational audienceWe describe an experimental platform for testing the DarkO tail-sitter drone in an open wind tunnel experiment. The DarkO convertible UAV is developed and 3D printed at the Ecole Nationale de l'Aviation Civile (ENAC), in Toulouse (France). The objective of the experimental platform is to allow testing control laws in a realistic and secure context. We propose a test bench with a single degree of freedom corresponding to the drone pitch axis. We design a linear proportional/integral feedback hovering stabilizer in the presence of constant wind, and we illustrate its effectiveness in stabilizing a hovering position through experimental results
Data-driven trajectory-based analysis and optimization of airport surface movement
No full textInternational audienceDriven by the recent deployment of equipment and the availability of data, aircraft trajectories on the surface can be tracked continuously, and the detailed analysis of surface operations is enabled to measure airport operational performance and prepare for collaborative decision making. This paper proposes a hybrid approach combining traffic analysis and optimization to address the airport surface movement problem. Important metrics of surface management such as runway utilization, taxi routes, taxi times are extracted and characterized based on Automatic Dependent Surveillance-Broadcast (ADS-B) messages. After recognizing the congestion bottleneck, a trajectory-based optimization model is proposed, and an adapted simulated annealing heuristic is applied to solve the related problem. Two concepts for optimizing surface operations: arrival taxi reroute and departure controlled pushback are evaluated. The proposed approach is illustrated for Beijing Capital International airport. Computational experiments show that a mean taxi-in time reduction of 5.1 min is achieved for the rerouting arrivals, and the mean taxi-out time is reduced by 3.7 min with gate holding strategy. This data-driven approach allows advanced characterization of complex surface operations and enables collaborative decision assistance tools for managing surface movement efficiently
Optimisation à grande échelle de trajectoires 4D d'avion
No full textAirspace capacity has become a critical resource for air transportation. The air traffic control workload is a significant factor leading to capacity limits of the air traffic management systems. Complexity in air traffic structure can lead to airspace saturation before reaching the capacity threshold. This issue motivates the development of decision-making tools to reduce airspace congestion while improving the air traffic controller’s situation awareness.This thesis presents an optimization approach to address the strategic decongestion planning problem in a trajectory-based operation environment. The congestion mitigation strategy relies on departure time adjustment, traffic re-routing, and flight level allocation methods. The objective function is developed from a congestion metric based on linear dynamical systems. The numerical properties of dynamical systems can quantify various traffic situations in the context of air traffic management. Further, the preceding problem is also extended by proposing the robust strategic decongestion planning model, where time uncertainties are considered to improve the robustness of the solution.The resolution methods based on metaheuristic and reinforcement learning-based hyper-heuristic approaches have been developed to solve the proposed large-scale problems. These methods are implemented and validated with real traffic data in the French airspace, involving more than 8,000 trajectories. The proposed resolution methods efficiently reduce the total congestion between trajectories at the strategic level. Further, the performance of the proposed hyper-heuristic outperforms different algorithms in terms of the remaining congestion.Finally, a comparative study between the strategic deconfliction and the strategic decongestion methods is carried out to compare the robustness of their solutions against departure time perturbation. The Monte Carlo simulation method is proposed to evaluate the robustness of the two solutions. The simulation results show that the strategic decongestion method is more robust against departure time perturbation than the strategic deconfliction method.La capacité de l'espace aérien est devenue une ressource critique pour le transport aérien. La charge de travail du contrôle aérien est un facteur important qui limite la capacité des systèmes de gestion du trafic aérien. La complexité de la structure du trafic peut conduire à la saturation de l'espace aérien. Afin de répondre à cette problématique, il est nécessaire de développer des d'outils d'aide à la décision pour réduire la congestion de l'espace aérien tout en améliorant la conscience de la situation du contrôleur aérien.Cette thèse présente une approche d'optimisation permettant d'aborder le problème de planification stratégique des trajectoires d'avions. Le but de cette approche est de réduire la congestion de l'espace aérien par modification des créneaux de décollage, des routes et des niveaux des vols. La fonction objectif est développée à partir d'une métrique de congestion basée sur des systèmes dynamiques linéaires. Les propriétés numériques des systèmes dynamiques permettent de quantifier les diverses situations du trafic dans le contexte de la gestion du trafic aérien. Le problème est également abordé en prenant en compte les incertitudes en proposant un modèle de planification stratégique robuste.Nous proposons deux méthodes de résolution basée sur des algorithmes de type métaheuristique et hyper-heuristique basées sur l'apprentissage par renforcement pour résoudre ces problèmes à grande échelle. Les simulations sont conduites sur l'espace aérien français, impliquant plus de 8,000 vols. Les méthodes de résolution proposées nous permettent de réduire efficacement la congestion entre les trajectoires au niveau stratégique. La performance de l'hyper-heuristique proposée surpasse celle de différents algorithmes en termes de congestion restante.Enfin, une étude comparative entre les méthodes de déconfliction stratégique et de décongestion stratégique est réalisée pour comparer la robustesse de leurs solutions face à des perturbations des temps de départ. Nous proposons une méthode de simulation de Monte-Carlo pour évaluer la robustesse des deux solutions. Les résultats de la simulation montrent que la méthode de décongestion stratégique est plus robuste face aux perturbations des temps de départ en comparaison de la méthode de déconfliction stratégique
Metropolis II: Centralised and strategical separation management of UAS in urban environment
No full textInternational audienceThis paper presents a centralized and strategical approach for Unmanned aircraft systems Traffic Management (UTM) to design optimal 4D trajectories minimizing the total flight time of all vehicles over a given time window. Potential losses of pairwise separation between vehicles are modeled and solved. A 4D trajectory is modeled by choosing an horizontal path (with an associated nominal speed profile), a departure slot and a cruising flight level. The problem is formulated as a mixed-integer linear program. A two-step solution approach is proposed that takes into account operational requirements, such as late flight intention deposits, or static and dynamic geofences; and that is able to deal with very high traffic density (up to 6300 vehicles in an horizon of one hour). Experimental results show that it is possible, by delaying flights at the departure or modifying their 4D route (vertically or horizontally), to obtain Unmanned Aircraft Systems (UAS) flyable trajectories that avoid losses of separation and minimize the total flown time
SSW-2D : un logiciel open-source et gratuit pour la propagation par méthode split-step wavelet
No full textInternational audienceDans ce papier, le logiciel SSW-2D pour la modélisation de la propagation par méthode split-step wavelet (SSW) est présenté. SSW améliore la méthode split-step Fourier (SSF) classiquement utilisée pour la modélisation déterministe de la propagation sur de grandes scènes. Cependant, coder cette méthode s'avère bien plus complexe et nécessite une compréhension fine de la décomposition en ondelettes. C'est pourquoi un logiciel clé en main s'avère utile pour permettre à la communauté de la propagation de s'approprier SSW. Ce logiciel open source (sous licence GNU GPL V3) vient avec une interface graphique pour une utilisation basique. Il est aussi possible de passer par des fichiers csv. Cette deuxième option est préférable si l'on veut modifier ou améliorer le code. Ce papier présente succinctement SSW et positionne cette méthode face à SSF. De plus, les modules composant le logiciel SSW-2D sont détaillés
Formal Verification of Graphical Properties of Interactive Systems
No full textInternational audienceCritical systems, particularly aeronautical systems, contain newly highly interactive devices: for example, the new generation cockpits use sophisticated electronics. They are driven by complex reactive software applications able to react to various kinds of inputs and to provide a representation of their internal state. In this context, the certification processes described in DO-178C and in DO-333 give an important place to formal verification of the requirements of these systems. Many formal methods have been proposed for this verification. However properties related to the graphical elements of these systems like position, overlapping, color, etc. have not received the same attention as others like safety, liveness, reachability or boundary ones. In this paper, we propose an original approach based on deductive verification to check graphically oriented requirements. On the basis of a semantic of reactive applications based on the weakest precondition calculus, we propose an algorithm for the verification of graphical requirements. This algorithm is developed in the context of Smala/Djnn: an environment for developing interactive systems. We illustrate our approach on the Traffic alert and Collision Avoidance System (TCAS), an aeronautical case study, and some of its graphical requirements
CONFLICT RESOLUTION WITH TIME CONSTRAINTS IN THE TERMINAL MANEUVERING AREA USING A DISTRIBUTED Q-LEARNING ALGORITHM
No full textInternational audienceWith the growing number of flights, more and more conflicts have to be solved in Terminal Manoeuvring Areas (TMAs). In order to keep a fluid flow of aircraft arriving on an airport, air traffic controllers use softwares to help them to solve conflicts and sequence aircraft on runways. This paper faces the sequencing and merging problem using a reinforcement learning algorithm (Q-Learning) in order to measure its performance. This algorithm has been run on a scenario representing a regular day at Paris Charles de Gaulle airport (CDG), and gives satisfying results. Then, it has been benchmarked on heavilyloaded scenarios, with more aircraft than the previous ones in order to see the limits of reinforcement learning efficiency. The Q-Learning algorithm can not only solve conflicts on this heavilyloaded scenario but it also has a reasonable computational time. By using a Q-learning algorithm in a distributed way, we aim to find an optimized solution on heavily-loaded scenarios without compromising the computational time
Ouroboros An efficient and provably secure KEM family
No full textInternational audienceIn this paper we introduce Ouroboros, a new family of Key Exchange protocols based on coding theory. The protocols propose a middle ground between the cryptosystems based on QC-MDPC codes, which feature small parameter sizes, but have a security reduction to two problems: the syndrome decoding problem and the indistinguishability of the code, and the HQC protocol, which features bigger parameters but has a security reduction to the syndrome decoding problem only. Ouroboros features a reduction to the syndrome decoding problem with only a small overhead compared to the QC-MDPC based cryptosystems. The approach is based on an ideal structure and also works for the rank metric. This yields a simple, secure and efficient approach for key exchange, the Ouroboros family of protocols. For the Hamming metric we obtain the same type of parameters (and almost the same simple decoding) as for MDPC based cryptosystems, but with a security reduction to decoding random quasi-cyclic codes in the Random Oracle Model. This represents a reduction of up to 38% on the public key size compared to HQC, for the most secure parameters. For the rank metric, we obtain better parameters than for RQC, saving up to 31% on the public key for the most secure set of parameters, using non homogeneous errors in Ouroboros. In this full version, the protocol and decoding algorithm have been slightly improved, additional details are given in the security proof, and the protocol is fully described for the rank metri
A new multi-commodity flow model to optimize the robustness of the Gate Allocation Problem
No full textInternational audienceThe Gate Allocation Problem (GAP) is one of the numerous operational problems that all busy airports have to handle and to optimize every day. In this article, we focus on two important objectives for the GAP: the robustness and the taxiing times. We first analyze a basic model which optimizes the robustness of the solution to the GAP and propose a more realistic and compact Multi-commodity Flow Problem model (MFP) to optimize both the robustness and the taxiing times. Based on tests with real data at Paris-Charles-de-Gaulle airport, this new efficient MFP model outperforms the basic one by orders of magnitude
Characterization of Real Multi-Correlator Outputs for SQM Performance Evaluation
No full textInternational audienceSignal Quality Monitoring is a process put in place in augmentation systems such as SBAS or GBAS to monitor potential signal distortions with high integrity that may be created by a satellite failure. It generally consists in the combination of several correlator outputs in so-called metrics, such as the single ratio metrics, the symmetric ratio metric or the double different metrics. To validate the compliance of a particular combination of metrics, it is necessary to validate the detection performance of an SQM process against every possible distortions of a Threat Space, in presence of typical errors affecting the metrics.Usually, theoretical models are used in order to simulate the error affecting the correlator outputs and the metrics. However, those models cannot fully capture the diversity of the errors, such as the temporal correlation of multipath, or its effects on close correlator outputs. It is therefore of high interest to use real data collect in order to derive the models of the correlator output models, to validate the compliance of an SQM in operational conditions.ENAC has put in place an automated data collect in order to observe the distribution of correlator output errors over a long period. Due to the large variation of the number of low-elevation satellites in a day, this scheduling task requires a specific process to collect as many observations as possible from low-elevation satellites in a limited period of time. An optimization algorithm, adapted from the simulated annealing process, allows to find an optimal scheduling, taking into account the constraint of the long post-processing task of the collected digitized samples by a software receiver.By accumulating a large set of correlator outputs from low-elevation satellites, an accurate distribution of the covariance matrix of the correlator outputs is obtained, capturing all the effects occurring in the real world and in a real receiver. Applying this distribution in the SQM compliance test can help to have a more realistic performance. The comparison of an SQM performance between theoretical and observation-based models shows some major differences