Archive ouverte de l'ENAC
Not a member yet
3458 research outputs found
Sort by
Imaginer des scenarios pour la mobilité écologique sur une île à l'horizon 2050
No full textIn the light of the current environmental crisis, it is crucial to imagine a new view of mobility. Thus, building representations of the future of mobility is needed to get a better insight into the trajectories to our environmental goals: this is possible with several approaches. This study aims to imagine credible and desirable futures, regarding both technological aspects and usage. Our approach describes the mobilities to and within an island in both quantitative and a qualitative aspect, regarding mobilities as both the consequence of the island’s characteristics and the structuring element of the residents’ living environment. We applied a systemic modelling methodology to a mobility case. This led to the determination of the model’s main dimensions, and the creation of four different scenarios to explore them. Implementing these scenarios on the software Magnitude enabled us to shed light on other questions related to mobility, such as the island’s resources, as well as its economic activity. We then discussed whether the selected methodology is suitable for the case of mobility and what it brought to the study.Dans un contexte de crise environnementale, il est crucial d’imaginer une nouvelle vision de la mobilité. Il est ainsi nécessaire de construire des représentations du future de la mobilité, afin d’adopter une meilleure perspective concernant nos objectifs environnementaux : plusieurs approches le permettent. Cette étude vise à imaginer des futures crédibles et désirables, autant sous l’aspect technologique et l’aspect de l’usage. Notre approche décrit les mobilités vers et dans une île de manière quantitative et qualitative, en plaçant les mobilités comme la conséquence des caractéristiques de l’île, mais aussi comme la structure du cadre de vie des résidents. Nous avons appliqué une méthodologie de modélisation systémique au cas des mobilités. Cela nous a permis de déterminer les dimensions principales du modèle, et de créer quatre différents scénarios pour les explorer. L’implémentation de ces scénarios dans le logiciel Magnitude a permis de soulever d’autres questions liées à la mobilité, telles que les ressources de l’île et son activité économique. Nous afin ensuite discuté de l’adaptation de la méthode utilisée dans le cas particulier de la mobilité, ainsi que son apport dans le cadre de notre étude
Multi Criteria Methodology for Aircraft Trajectory Planning Algorithm Selection: A Survey
No full textInternational audienceAircraft trajectory is one of the most fundamental objects in air traffic management. Its optimization is essential to ensure efficient and sustainable aviation. This survey proposes to study all the phases of a flight, from its prediction several days before day of flight to the landing of the aircraft, including also the study of a possible emergency situation. Each phase of flight raises different issues and is subject to particular constraints. These guide the choice of potentially usable optimization methods. This study proposes, from the context, the issues, and existing studies, a methodology to identify the most appropriate solution algorithms for optimizing each phase of flight. This methodology is based on 5 evaluation criteria: optimality, computing time, adaptability, memory usage, and multi-trajectories. Finally, thanks to it, some methods are compared based on their consistency with solving problem associated to each phase of flight
UASOS: An Experimental Environment for Assessing Mental Fatigue & Cognitive Flexibility during Drone Operations
No full textInternational audienceMental fatigue from continuous operations without breaks represents a safety issue for military drone operations, as these systems are complex and operate during long shifts. Military operations are hard to study due to their sensitive nature. The open-access program UASOS serves as a testbed to examine the effects of mental fatigue in an ecologically valid environment. UASOS recreates fundamental aspects of military drone operations in a controllable environment that is easy enough for novices to understand but demanding enough to elicit mental fatigue. Participants alternate between navigating a drone-using either a trackball/mouse or a joystick-and searching for visual targets. The protocol is set up in a way that taxes the cognitive flexibility of participants by constantly requiring them to alternate between tasks. In addition, several parameters such as difficulty, duration, questionnaires, training phases, and more can be adapted. The task also allows for synchronization with physiological data using LabStreamingLayer. Implemented in python, the code is set up to be easily installed
BOUNDARY LAYER TRIPPING STRATEGIES FOR SUPERSONIC AIR INLET APPLICATIONS
No full textInternational audienceThis paper presents a numerical study for different boundary layer tripping solutions on a simplified geometry of a compression ramp found in supersonic air inlet configurations. The investigated tripping solutions will force the boundary layer to be turbulent at the tip of the external supersonic diffuser, where the first shock appears. The different proposed solutions consist of a geometrical trip, a bypass transition with an isotropic homogeneous turbulence injection in the domain and a combination of both. We perform a characterization of the boundary layer developing on the simplified ramp configuration and compare the different tripping strategies.</div
Contrôle actif d'un micro-drone convertible face à la turbulence
No full textDrones have become a prevalent tool in numerous fields, including inspection, surveillance, and maintenance. However, one area where they are currently lacking is autonomy. Tailsitter offer a viable solution, combining large flight envelopes with energy efficiency. Nonetheless, tailsitter unmanned aerial vehicles are particularly susceptible to disturbances, specifically turbulence during hover phases. This is because of the large vertical surface area of the wing, which offers a high degree of wind resistance. Furthermore, the drone's body rotates during transition, making it difficult to accurately measure airspeed. It is not possible to accurately measure wind speed, whether the drone is hovering or moving at a low speed. Additionally, this underactuated drone, has two motors on the wing and two aerodynamic surfaces on the trailing edge. The interaction of the airflow from the propellers with the elevons results in a coupling between the actuators. The aim of this thesis is to examine the control of drone in challenging environments, with a particular emphasis on the impact of wind. However, initial research focused on windless dynamics in order to gain a better understanding of the simplified dynamics. We were able to propose a non-linear modification of the control vector to transform this nonlinear model into a input-affine model. Based on this model, we proposed a local-global control law based on hybrid dynamics with hysteresis, which allowed us to extend the stability domain of an aggressive linear control law by means of a non-linear law with a large region of attraction, but which was less aggressive. Further work was conducted on the stabilization of a tailsitter drone under wind steps. This resulted in a characterization of stationary equilibria for a range of wind conditions and a linearized representation of the drones dynamics. Using this model, we were able to analyze the saturation levels of actuators and the authority available close to the equilibrium points. We have implemented a stabilization system based on output feedback with proportional and integral action, derived from the model. This control does not utilize the drones pitch angle measurement, since the target value is not known a priori. It would require an estimation of wind speed and direction, which is not feasible. The loop is optimized using "Systune" software in order to achieve effective disturbance rejection properties. We adopted an incremental approach by initially evaluating the control law on a one-degree-of-freedom model against an open wind tunnel. Upon validation, we proceeded to implement the control law within the Paparazzi UAV system. Due to its modular design, we were able to establish a connection with the state estimation and actuators, enabling us to execute flights on the entire model with six degrees of freedom. We then proposed an architecture inspired by the tailsitter, called freewing. We have developed a multi-body drone that is based on a wing that can freely rotate on its pitch axis around a fuselage. The wing is driven the same way as the tailsitter, and the fuselage has a motor and an aerodynamic surface to keep it horizontal. With this architecture, we aim to achieve a natural passivity to turbulence induced by the natural change in wing incidence as a function of the incident wind. However, there is also the possibility of installing a payload on the horizontal fuselage. To obtain a simulation model, we have modeled the drone's dynamics using the Udwadia-Kalaba equations and the phi-theory. We focused on stabilization and guidance of the UAV through the use of incremental nonlinear dynamic inversion (INDI). The wing and fuselage actuators are used to achieve a global stabilization law. In order to evaluate the benefits of this architecture, we conducted flight tests.Les drones sont devenus un outil dans de nombreux domaines tels que l'inspection, la surveillance ou la maintenance. Cependant, ils souffrent d'une autonomie limitée. Les tailsitters apportent une solution grâce à une grande enveloppe de vol et à une efficacité énergétique. Toutefois, les tailsitters sont grandement sujets aux perturbations aérologiques et notamment aux turbulences dans les phases stationnaires principalement. Cela est dû à la grande surface d'aile verticale, laquelle possède une grande prise au vent. De plus, leur corps tournant lors de la transition, il est donc compliqué de mesurer la vitesse de l'air. Ainsi, en stationnaire ou à faible vitesse, le vent n'est pas connu. Ce type de drone est sous-actionné puisque l'on trouve deux moteurs sur l'aile et deux surfaces aérodynamiques sur le bord de fuite. Le flux d'air des hélices soufflant les élevons, nous avons un couplage entre les actionneurs. Cette thèse cherche à étudier la commande de drones dans des environnements perturbés ou en présence de vent. Les premiers travaux se sont concentrés sur la dynamique sans vent pour appréhender une dynamique simplifiée. Nous avons pu proposer une modification non-linéaire du vecteur de commande pour rendre ce modèle linéaire en commande. De ce modèle, nous avons proposé une loi de commande locale-globale fondée sur une dynamique hybride à hystérésis. Elle permet d'étendre le domaine de stabilité de la loi de commande linéaire agressive à l'aide d'une loi non-linéaire avec une grande région d'attraction, mais moins agressive. La suite des travaux s'est concentrée sur la stabilisation d'un tailsitters soumis à des échelons de vent. Il en résulte une caractérisation des équilibres stationnaires pour un ensemble de conditions de vent et l'obtention de la représentation linéarisée de la dynamique du drone. À l'aide de ce modèle, il a été possible d'analyser les saturations des actionneurs et l'autorité disponible aux environs des points d'équilibre. Nous avons réalisé une stabilisation établie sur un retour de sortie, avec une action proportionnelle et intégrale. Cette commande n'utilise pas la mesure de l'angle de tangage du drone, car nous ne pouvons pas, a priori, connaître la valeur cible qui nécessiterait une estimation de la vitesse et de la direction du vent. L'optimisation de ce bouclage est effectuée à l'aide du logiciel "Systune" pour obtenir de bonnes propriétés de réjection de perturbation. Une approche incrémentale a été suivie, la loi de commande ayant été testée dans un premier temps sur une maquette à un degré de liberté face à une soufflerie ouverte. Une fois validée, la loi de commande a été implémentée dans le système de drone Paparazzi. Grâce à son architecture modulaire, il a été possible de nous interfacer avec les codes d'estimation et de commande des actionneurs. Ainsi, nous avons pu réaliser des vols sur le modèle complet à six degrés de liberté. Enfin, nous avons proposé une architecture inspirée du tailsitter, nommée freewing. Nous avons développé un drone multicorps basé sur une aile en rotation libre sur son axe de tangage autour d'un fuselage. L'actionnement de l'aile est sensiblement le même que pour le tailsitters et le fuselage possède deux actionneurs pour se maintenir horizontal. Nous recherchons, dans cette architecture, une passivité naturelle à la turbulence induite par le changement naturel de l'incidence de l'aile en fonction du vent incident. Il s'agit aussi d'installer une charge utile sur le fuselage horizontal sur le domaine de vol. De plus, nous avons réalisé un modèle de simulation où la dynamique est obtenue à l'aide des équations de Udwadia-Kalaba et de la phi-théorie. Enfin, nous nous sommes concentrés sur la stabilisation et le guidage du drone en utilisant une inversion incrémentale non-linéaire de la dynamique (INDI). Nous utilisons les actionneurs de l'aile et du fuselage pour obtenir une loi de stabilisation globale. Des vols ont validé l'intérêt de cette architecture
Une critique de l’hylémorphisme dans les théories du design d’interaction : recherche de concepts dans la philosophie pour appréhender la forme et la matérialité
No full textInternational audienceLa recherche d’une compréhension de l’émergence d’un design sur le plan de la forme et de la matière est ancienne. On observe une montée en puissance du concept de forme dans la littérature scientifique sur le design d’interaction, en parallèle d’un tournant matériel, pour exprimer les forces à l’œuvre dans la conception. Dans cet article, nous soulignons en quoi les théories actuelles nous semblent implicitement construites sur le paradigme hylémorphique, et ce, de manière peu cohérente. Pour rendre compte plus finement des mécanismes d’élaboration et de conception dans le design d’interaction, nous proposons d’utiliser les concepts explicites de transduction, de métastabilité, d’individuation et de concrétisation, tirés de l’approche critique du schéma hylémorphique de Gilbert Simondon. Au paradigme hylémorphique de la mise en forme aristotélicienne, nous opposons ainsi celui, transductif, de la prise de forme. Pour une meilleure compréhension, nous présentons une illustration de ces concepts dans un exemple : un dispositif tactile à changement de forme pour les pilotes d’avion de ligne
Past, current and future trend for the usage of extend reality (XR) in aviation
No full textInternational audienceExtended, Virtual, Mixed and Augmented reality (XR, VR, MR, AR) devices, allow users to interact with computer-generated simulations that mimic or enhance the perception of the real world. AR overlays digital information onto the actual environment, while VR immerses users entirely in a virtual environment. MR combines elements of both to seamlessly blend the virtual and real worlds. XR technology provides immersive experiences, creating a sense of presence in the virtual realm. XR applications span various domains, such as training, education, medical, digital health, and aviation, benefiting pilots and air traffic controllers (ATCOs). In this chapter, we share our experiences using XR technology, detailing findings and the process of efficient implementation for end users like pilots and ATCOs. We present examples of improving the paper strip by projecting dynamic and updated information, ensuring smoother monitoring for controllers. Our remote tower project incorporates remote sensing with sound spatialization, enhancing situation awareness during airfield management. We also discuss developing an immersive analytic tool using XR devices for investigating recorded aircraft trajectories and its potential replication through dedicated APIs. We explore prospects for studying human behavior using VR devices under more immersive conditions, manipulating mental workload, prototyping and evaluating complex interfaces (e.g., cockpit design), and understanding conflict resolution algorithms through eXplainable AI (XAI). Finally, we address future challenges, such as improving interaction capabilities with tangibility in VR, and the opportunities XR technology could bring to aviation activities
Visual alerts for operator cognitive flexibility improvement: a neuroergonomic approach
No full textInternational audienceThis study investigates the mitigation of cognitive flexibility decrements during long-endurance Uncrewed Aerial Systems (UAS) missions through visual alerts. UAS mishaps are likely partially attributable to decreased cognitive flexibility caused by mental fatigue. Two groups of participants performed a 2-hour UAS simulation. One group was supported by visual alerts when switching between tasks, while the other was not. To further investigate the effect of visual alerts, the participant’s cerebral and cardiac activity during the task was recorded using electroencephalography (EEG) and electrocardiography (ECG). Results showed a weaker subjective increase of mental fatigue across time with visual alerts than the control group. Furthermore, behavioural results showed improved reaction time and accuracy in some performance metrics. Future work may improve this system’s efficacy by implementing it with an adaptive interface
Air traffic classification using convolutional neural networks
No full textInternational audienceThe difficulty of managing airspace is reflected in the complexity of forecasting its evolution. This paper presents anew neural network framework for managing images for which pixels are matrices with application to air trafficcomplexity map prediction. By modelling air traffic with a linear dynamical system, air traffic maps can be definedas images whose pixels are matrices. By computing intermediate steps, these air traffic maps are defined as imageswhose pixels are symmetric positive definite matrices. Then, we implement a convolution neural network with aspecific data preprocessing step, new convolution, max-pooling, and flatten layers suitable to such images. The newconvolution, max-pooling and flatten layers are capable of processing images coming from the data preprocessingstep
Cartan moving frames and the data manifolds
No full textInternational audienceThe purpose of this paper is to employ the language of Cartan moving frames to study the geometry of the data manifolds and its Riemannian structure, via the data information metric and its curvature at data points. Using this framework and through experiments, explanations on the response of a neural network are given by pointing out the output classes that are easily reachable from a given input. This emphasizes how the proposed mathematical relationship between the output of the network and the geometry of its inputs can be exploited as an explainable artificial intelligence tool