Archive ouverte de l'ENAC
Not a member yet
3458 research outputs found
Sort by
Wasserstein distance based semi-supervised manifold learning for GNSS multipath detection
No full textNational audienceOne of the main weaknesses of the Global Navigation Satellite System (GNSS) is the sensitivity to multi-path disturbances. A multi-path is a reflection of the signal emitted by a navigation satellite on one or more surfaces surrounding the receiving antenna. These multi-paths degrade the precision and accuracy of the position calculated by the GNSS receiver. This could lead to potential hazard especially in safety of life applications. Aircraft navigation which heavily depends on GNSS nowadays is a typical illustration. Thus, detecting the possible distortion in the input data is of prior importance. Previous studies have shown the interest of supervised Machine Learning (ML) techniques in the purpose of detecting distortion in GNSS signals. However, real labelled data are scarce. This is a genuine difficulty as supervised ML algorithms require in general large amount of labelled data for training.The aim of this research work is to consider semi-supervised learning as a mean to improve the performance of a fully supervised model by means of unlabelled samples added to the training dataset.The underlying principle is that the manifold of the data can also be learned by the network through unlabelled samples if a suitable measure of similarity with labelled samples is provided. The metric chosen in this study is the Wasserstein distance also known as the Earth Mover's Distance (EMD).The implementation of the unsupervised contribution to the training process takes the form of a regularisation term in addition to the classical supervised Binary Cross Entropy (BCE) loss function. This unsupervised term is linear in sample thus it preserves the ability to process large data sets by batches of data and it benefits from stochastic gradient based techniques during training too
Power Cost Comparison of Neural-Network Compression Methods for Satellite Imagery
No full textInternational audienc
Cognitive Incapacitation in Aviation
No full textInternational audienceAbstract: In aviation, cognitive incapacitation refers to a temporary reduction in a pilot’s cognitive performance, typically triggered by unexpected and stressful events. This situation can alter perception, reasoning, decision-making performance, and ultimately the ability to perform cockpit duties effectively. Cognitive incapacitation has been identified as an important factor in many aviation accidents and incidents. However, research is scarce, and varying terminology is used to describe the phenomenon. Additionally, the methods to effectively induce and measure cognitive incapacitation remain debated. We conducted a comprehensive scoping review to enhance our understanding of this phenomenon and gain clearer insights into the research protocols. This review may be applied for improving training and developing solutions to mitigate the effects of cognitive incapacitation
Simulations of longitudinal combustion instabilities in a LOX-CH4 Liquid Rocket Engine
No full textInternational audienceThe present paper studies combustion dynamics in a single injector, cryogenic CH4/O2 rocket combustor at subcritical pressures. Such conditions are usually encountered in deep-throttlable engines, needed for launcher reuse. Large Eddy Simulation (LES) is employed with a homogeneous multi-fluid approach and real gas thermodynamics to simulate the DLR 'BKN' experimental combustion chamber. After presenting the experiment and simulation setups, numerical results are compared with experimental data from stable and unstable regimes, providing valuable insights into the accuracy and limitations of the employed numerical models
Spherical and Plane Waves Propagation through a Realistic Turbulent Marine Layer
No full textInternational audienceIn this paper, the impact of a realistic turbulent medium obtained from fluid mechanics equations on radiowave propagation is studied. The realistic medium contains more structured variations of refractive index than the one usually generated from Gaussian-like spectra. The impact on propagation is illustrated by comparing the effect on sources presenting both a spherical and a plane wavefront. These results highlight the impact of the localised atmospheric structure that could reveal an underestimation of the turbulence impact when obtained via the commonly used Gaussian-like spectra.</div
DVSim : un simulateur de vision événementielle pour l'apprentissage de réseaux de neurones à impulsions
No full textInternational audienceDVSim : un simulateur de vision événementielle pour l'apprentissage de réseaux de neurones à impulsion
A Wavelet-Based Computation of the Field Received by a LEO Satellite in Radio Occultation
No full textInternational audienceThe propagation of an electromagnetic field from the edge of the atmosphere to a satellite in Low Elevation Orbit (LEO) involves the computation of a diffraction integral which evaluation can be computationally demanding when a direct numerical integration is performed. Previous studies have introduced efficient methods to perform this calculation, but these approaches often rely on physical assumptions that may compromise the accuracy of the results. In this paper, we present a novel approach that avoids such assumptions and efficiently computes the field received by the LEO satellite using a wavelet decomposition of the integral
Enhancing Airport Operations: An Integrated Framework for Optimising Aircraft Surface Movement and Airport Resource Management
No full textInternational audienceThe ever-increasing volume of air traffic has significantly strained airport resources, intensifying congestion and emphasising the critical interface between airspace, airside and landside operations in airport management. As a result, optimising aircraft surface movement and resource allocation becomes essential to improving airport efficiency within these constraints.This paper presents an integrated optimisation framework designed to coordinate airport surface traffic with resource management at the tactical level, enhancing operational efficiency and resilience while maintaining the balance between demand and airport capacity. Unlike traditional gate allocation methods that address departures and arrivals independently, this study emphasises the efficiency of assigning paired arrival and departure flights of the same aircraft to a single gate, reducing the costs and delays associated with towing aircraft between gates. A hierarchical allocation scheme is proposed to assess gate and terminal overload, accounting for aircraft in-block and off-block times, rather than assuming fixed schedules. Further, this study integrates ground handling assignments into airport resource management optimisation, introducing an additional level of uncertainty while capturing the complexities of daily airport operations.A novel integrated Mixed Integer Programming (MIP) model is developed to formulate the problem mathematically. To ensure a robust and flexible solution that aligns with real operational scenarios, a metaheuristic approach is incorporated to manage uncertainties in aircraft surface movement and resource availability. The algorithm employs a combined probability- and performance-based decision-making process to guide the neighbourhood operator. Additionally, a sliding window technique, with a 2-hour window length and a 30-minute shift, is applied to optimise the model and validate the decision dynamically.The proposed framework and algorithm are validated through a case study using historical radar data from Paris Charles de Gaulle (CDG) Airport, demonstrating its applicability and effectiveness in real-world applications
Strategic vs. altruistic Corporate Social Responsibility
No full textNational audienceThe concept of Corporate Social Responsibility (CSR) has evolved since Milton Friedman’s 1970 assertion that a business’s sole responsibility is profit. Today, global frameworks like the UN Global Compact and EU regulations emphasize corporate accountability, particularly regarding social and environmental impacts.Corporate Social Responsibility (CSR) has become central in discussions of firm behavior, governance, and public goods provision. CSR however varies across firms. Some adopt basic strategic CSR (b-CSR), considering social and environmental issues only to the extent that they affect consumer demand and profitability. Others practice environmentally committed CSR (e-CSR), internalizing the full social cost of emissions. A few pursue fully committed CSR (w-CSR), aiming to maximize overall social welfare. The paper analyzes CSR’s effects on firm behavior through economic modeling. Itfirst examines a single firm producing CO2 emissions, where reducing emissions increases costs but appeals to environmentally conscious consumers. Three firm types—b-CSR, e-CSR, and w-CSR—are considered.The study then extends to a competitive market with two firms engaged in Cournot competition. It examines scenarios where firms have different CSR commitments, analyzing how competition, emissions, and profits are affected. Finally, the paper compares these outcomes to an ideal scenario where firms are regulated to maximize social welfare