Archive ouverte de l'ENAC
Not a member yet
3458 research outputs found
Sort by
Bayesian EM Approach for GNSS Parameters of Interest Estimation Under Constant Modulus Interference
No full textThis version contains an erratum for Eq. (22çInternational audienceInterferences pose a significant risk to applications that rely on Global Navigation Satellite Systems (GNSS). They have the potential to degrade GNSS performance and even result in service disruptions. The most notable type of intentional interference is characterized by a constant modulus, such as chirp and tone interferences. These interferences have a straightforward structure, leading to the creation of complex circles when attempting to identify their contribution. To address the interference and improve the situation, we calculate the maximum likelihood estimator for the relevant parameters (time delay and Doppler shift) while considering the presence of these latent variables. To achieve this, we employ the Expectation Maximization algorithm, which has previously demonstrated its effectiveness in similar scenarios. Experiments conducted using synthetic signals confirm the efficiency of the proposed algorithm
Schedule optimization and staff allocation for Airport security checkpoints using Guided Simulated Annealing and Integer Linear Programming
No full textSecurity checkpoints are an important matter of concern for airport operators. When functioning effectively, they not only maintain the airport overall safety at a high level, but also provide passengers with a positive airport experience. The perceived quality of service at the airport greatly depends on the time spent by passengers at the security lines. To ensure optimal safety performance, screening lines have a limited capacity of passengers they can handle. Thus, to prevent extended waiting times for passengers, airports can only adjust the number of simultaneously open check lines. The airport operator must establish optimal schedules for opening security checkpoints and allocating necessary staff. Building upon a prior study focused on predicting the flow of passengers through the security checkpoints, this paper explores simulated annealing algorithm in conjunction with a queue simulator and an integer programming algorithm to establish the most effective opening schedule for security checkpoints based on the prediction given by this previous study. The presented approach also determines the best allocation for dedicated staff based on the forecasted passenger flow. This approach limits the number of open security lines and ensures a waiting time below the maximum limit of 45 minutes set by the airport. It also complies with the work regulations that security agents are subject to
Integrating rigorous qualitative methods into the design and evaluation of safety-critical systems
No full textInternational audienceThe traditional approach to studying aviation systems applies Human Factors (HF) methods to generate task models, specify relevant functionality and measure system and user performance. Typical experiments establish protocols that test whether or not users perform tasks correctly. However, this approach largely ignores problems that arise under real-world conditions, such as when overloaded pilots are confused and switch off cockpit instruments. We argue that aviation researchers can benefit from a mixed-method approach that combine rigorous qualitative HCI methods to assess design variants in realistic situations. For example, the comparative structured observation method presents users with directly comparable tasks with selected design variants and asks them to reflect deeply on the tradeoffs associated with each. These studies follow best practices from controlled-experiment design, e.g.counter-balancing for order, but emphasize the collection of rich user insights over quantitative performance measures. When properly controlled, qualitative studies offer an important complement to theoretical task models and quantitative experiments to better understand how users will interact with proposed system designs. This paper demonstrates how to incorporate rigorous qualitative methods, with examples, into the process of designing and evaluating safety-critical systems
Single Frequency GNSS Carrier Phase Cycle Slip Detection and Identification Using a Factor Graph Approach
No full textInternational audiencePrecise and robust GNSS positioning with low-cost receivers is a challenge, as these receivers will often be disturbed by biased observations due to a given environment, for example by code pseudoranges with multipath. One way to reach centimeter positioning accuracy is by using very precise observations, in particular by using GNSS carrier phase measurements. These carrier phase measurements can be very precise (under centimeter standard deviation) but they contain an unknown, the phase ambiguity. Throughout time, for a tracked satellite during multiple epochs, this phase ambiguity is supposed to remain constant. So by building an observation model containing observations from multiple epochs, we can take advantage of this information. This is what is usually done in a discrete-time Kalman filter: we estimate the ambiguity of each satellite, and we model this ambiguity to be constant in the state transition model. However, in practice, cycle slips happen. If we suppose that the ambiguity is constant, then the positioning algorithms solutions will be degraded when cycle slips happen. This work uses factor graph optimization in order to detect and identify cycle slips on GNSS carrier phase observations from a single frequency receiver. This is done in a two-step process. It starts with a theoretical and very general approach to detect faults on measurements, as a binary decision. This is done with a theoretical distribution of a residual-based statistic (difference between observation models and measurements), and a comparison of the observed statistic with its theoretical distribution. In the case of detected fault, then comes step two of the process: by building multiple models and an optimization problem over a finite and discrete set of hypotheses, we are able to identify the observation where a fault was introduced. Following this theoretical approach to detect and identify faults, factor graph optimization is introduced. This is the strength of our approach: we are able to apply the procedure of detection and identification of faults for any factor graph, i.e., for any set of observations and states to estimate. Finally, the detection method is implemented in a GNSS simulator. We build a factor graph that estimates a receiver position from multiple carrier phase measurements. We apply the detection method to detect cycle slips on phase measurements. We show that, depending on the duration of the observation window, our detection method works for very small cycle slips, of one wavelength. We also compare Monte-Carlo simulations against theoretical probabilities of detection
Compile-Time Optimization of the Energy Consumption of Numerical Computations
No full textInternational audienceOver the past decade, precision tuning has become one of the keytechniques for achieving significant gains in performance and energy efficiency. This process consists of substituting smaller datatypes to the original data types assigned to floating-point variablesin numerical programs in such a way that accuracy requirements remain fulfilled. In this article, we discuss the time and energy savingsachieved using our precision tuning tool, POPiX. We validate ourresults on a set of numerical benchmarks covering various fields
Towards conformal automation in air traffic control: Learning conflict resolution strategies through behavior cloning
No full textInternational audienceA critical factor in achieving conformity of automation tools in performing expert tasks, such as air traffic conflict resolution, is the identification of air traffic controllers’ (ATCOs’) preferences (conflict resolution strategies) and the automation tool’s ability to learn and recommend solutions that incorporate these preferences. We propose a machine learning-based framework to learn and predict ATCOs’ conflict resolution preferences through behavior cloning. This framework is an ensemble of five regressor and classifier models. The conflict resolution data to train the machine learning models was collected from 8 experienced enroute ATCOs. The prediction results demonstrate that the ATCOs’ strategies encoded in the data can be learned by the model with high accuracy for the classification tasks and with low mean absolute error (MAE) for the regression task (for instance, the classification accuracy of above 92.7% for predicting the maneuvering aircraft, MAE for maneuver initiation distance ¡ 5.3 NM, MAE for predicting the heading angle ¡ 5.3°) for the ATCOs’ datasets. A sensitivity analysis performed to test the model robustness demonstrates that the proposed models are robust to up to 7.5% added Gaussian noise (with a mean equal to the value of each feature and varying standard deviation) to the dataset. In addition, we discuss the extent of acceptance of these predictions by the ATCOs through an ATCO acceptance exercise involving two ATCOs who demonstrate different conflict resolution strategies. ATCO selected the original strategy as one of the resolution preferences for 97% of the scenarios and the predicted strategy as one of the options for 78% of the scenarios. ATCO selected the conflict resolution strategies depicting ATCO B’s original strategies for 68% of the scenarios. The results from the acceptance exercise demonstrate that the proposed machine learning model can generate ATCO conformal predictions. The presented results and discussions also demonstrate the viability of using behavior cloning with chained predictions to develop individual and group conformal automation assistance tools for ATCOs
Approches alternatives pour penser et construire le futur du transport aérien : exemples d'expériences pédagogiques.
No full textInternational audienceLes projets présentés ont pour but d’accompagner les étudiant.es sur l’exploration de questions, de directions ou de solutions visant à surmonter la crise environnementale. Nous proposerons aux participants de discuter les sujets des projets (e.g. intermodalité, transport dual passager-bagages, low-tech en aviation, propulsion vélique, etc …) ou d’en proposer d’autres, et de réfléchir aux méthodes appropriées (e.g réflexivité, créativité, design spéculatif, modélisation, etc…)
Hidden Markov Models and Flight Phase Identification
No full textInternational audienceThe use of Hidden Markov Models (HMMs) in segmenting flight phases is a compelling approach with significant implications for aviation and aerospace research. It leverages the temporal sequences of flight data to delineate various phases of an aircraft's journey, making it a valuable tool for enhancing the analysis of flight performance and safety. In this work, we implement a multivariate HMM to identify 6 flight phases: taxi, takeoff, climb, cruise, approach and rollout. We reach a median global accuracy of about 97% over a sample of several thousand flights with a very low number of decoded unlikely transitions. Regarding several performance metrics, our method is competitive with existing methods in the literature, such as fuzzy logic. Additionally, it provides, for each point of the flight, a probability of belonging to each phase. Even in situations where there are missing values in the data, HMMs remain effective, ensuring that no critical information is lost during the segmentation process. We show that HMMs work seamlessly with the fine granularity of Flight Data Recorder (FDR) data. HMMs offer remarkable flexibility and adaptability, proving particularly effective when the number or order of phases is unknown or not predetermined, as is often the case with complex flight scenarios such as helicopter flights. This adaptability is crucial for handling the diverse range of flight operations that differ from one aircraft to another. An example is given with the segmentation of an Automatic Dependent Surveillance-Broadcast (ADS-B) helicopter flight operated by the Swedish National Police
Asymptotic efficiency for Sobol' and Cramér-von Mises indices under two designs of experiments
No full textA variety of indices aim to quantify the impact of input variables on a response, typically the output from a complex computer code or black-box model. Most commonly used, the Sobol' index typically measures the influence of some inputs from an explained variance perspective. However, some situations may require a more targeted analysis of some inputs influence. With no prior information, distribution-based measures appear to be appealing. In this purpose, so-called Cramér-von Mises indices (and their generalization) have been proposed in the literature, defined as an excess probability integrated over the output distribution that aim to reflect influence on the whole distribution of the output rather than on the variance solely. Inference of these various indices has remained a challenging topic especially in presence of many inputs. While several Sobol' indices estimators are known to be optimal under regularity conditions, the issue of asymptotic efficiency for Cramér-von Mises indiceshas been unaddressed in the literature so far. For these indices, we derive in this paper the efficiency bounds and discuss the known methods to achieve such optimal bounds. Two estimation contexts are considered: the so-called Pick-Freeze scheme and the Given-Data setting, for which the estimation is produced from a unique input-output sample
