Artificial Intelligence Solutions to Meteo-Based DCB Imbalances for Network Operations Planning

Livrables

This deliverable define the target operational framework (requirements, scenarios and use cases) for the evaluation of effectiveness of ISOBAR solution.

Final Project Results Report

This deliverable provides the ISOBAR findings which will include the description of final ISOBAR Solution a summary of proposed concepts and engines and the results with evidence of benefits and operational feasibility Furthermore technological risk and costs are analysed and regulatory implications are proposed together with a preliminary plan for the next RD phases

Multi-model probability of convection on a set of use cases

This deliverable provides probability of convection for a set of use cases focusing on the local and regional ATFCM needs

Applicable PF and evaluation reference

This deliverable provides a description of the ATFCM Performance framework and the reference for evaluation corresponding to the identified use cases

ML demand prediction model

This deliverable describes a function capable of estimating the demand fluctuations due to convective weather The deliverable will be updated at T016 to integrate further refinement of the model

Report on ISOBAR Evaluation and roadmap for ISOBAR B2B service

This deliverable describes the performance of ISOBAR solution against the baseline reference It includes the execution of the simulation campaign which covers the execution of ISOBAR solution service prototype to produce mitigation strategies tailored to scenarios for the selected use cases The report will elaborate recommendations for the next RD steps defining a highlevel roadmap for further development of ISOBAR service and integration in NM B2B services

Enhanced ATFCM Process and Service Requirements

Integrated operational flow of ISOBAR processes and models that will guide the developments in WP2 WP3 and WP4 The deliverable will be updated at T022 to integrate the results of technical WPs including feedback from the evaluation of effectiveness and consolidate them into an enhanced ATFCM process incorporating convective weather information The deliverable update will also include final requirements from T13

ISOBAR prototype and HMI showcase

This deliverable provides an experimental prototype with all ISOBAR modules integrated in and usable for evaluation through simulations The final HMI would be standalone designed to understand the concept and ease the dissemination of the project

Enhanced DCB algorithm with reinforcement learning

This deliverable provides an optimisation model for an enhanced DCB process The optimal coefficients will be computed using reinforcement learning based on the feedback in tactical ATFCM including an assessment of the relevant KPIs for each set of DCB solutions

Storm predictive model

This deliverable addresses the development of an enhanced convection indicator capable of identifying locationseverity and time window of storms

Hotspot detection library, based on demand and capacity characterisation

This deliverable will develop a library to identify hotspots in the airspace system given the demand and capacity profiles developed and refined

Publications

Integrated Frameworks of Unsupervised, Supervised and Reinforcement Learning for Solving Air Traffic Flow Management Problem

Auteurs: C. Huang and Y. Xu
Publié dans: 2021 IEEE/AIAA 40th Digital Avionics Systems Conference (DASC), 2021, Page(s) pp. 1-10
Éditeur: IEEE
DOI: 10.1109/dasc52595.2021.9594397

Multi-Agent Deep Reinforcement Learning for Solving Large-scale Air Traffic Flow Management Problem: A Time-Step Sequential Decision Approach

Auteurs: Y. Tang and Y. Xu
Publié dans: 2021 IEEE/AIAA 40th Digital Avionics Systems Conference (DASC), 2021, Page(s) pp. 1-10
Éditeur: IEEE
DOI: 10.1109/dasc52595.2021.9594329

Simulated-Annealing Hyper-Heuristic for Demand-Capacity Balancing in Air Traffic Flow Management

Auteurs: Khassiba, A., and Delahaye, D.
Publié dans: 12th SESAR Innovation Days, Numéro 5-8 December SIDs 2022, 2022
Éditeur: SESAR JU

Predicting Convective Storm Characteristics using Machine Learning from Hi-Resolution NWP Forecasts

Auteurs: Aniel Jardines, Manuel Soler, Javier García-Heras, Matteo Ponzano, Laure Raynaud, Lucie Rottner, Juan Simarro, and Florenci Rey
Publié dans: Conference Paper published at the General Assembly of the European Geoscience Union (EGU’21), Numéro 19–30 Apr 2021, EGU21-7516, 2021
Éditeur: EGU General Assembly 2021
DOI: 10.5194/egusphere-egu21-7516

Optimal Air Traffic Flow Management Regulations Scheme with Adaptive Large Neighbourhood Search

Auteurs: Ramón Dalmau, Gilles Gawinowski & Camille Anoraud
Publié dans: 12th SESAR Innovation Days, Numéro 5-8 December SIDs 2022, 2022
Éditeur: SESAR JU

Convection indicator for pre-tactical air traffic flow management using neural networks

Auteurs: Aniel Jardines; Manuel Soler; Alejandro Cervantes; Javier García-Heras; Juan Simarro
Publié dans: Machine Learning with Applications, Numéro Machine Learning with Applications 5 (2021) 100053 (https://www.sciencedirect.com/science/article/pii/S2666827021000256?via%3Dihub), 2021, ISSN 2666-8270
Éditeur: Elsevier
DOI: 10.1016/j.mlwa.2021.100053

Locally generalised multi-agent reinforcement learning for demand and capacity balancing with customised neural networks

Auteurs: Yutong Chen; Minghua Hu; Yan Xu; Lei Yang
Publié dans: Chinese Journal of Aeronautics, Numéro 24 January 2023, 2023, ISSN 1000-9361
Éditeur: Press of Acta Aeronautica et Astronautica Sinica
DOI: 10.1016/j.cja.2023.01.010

Comparison of various temporal air traffic flow management models in critical scenarios

Auteurs: Ramon Dalmau; Gilles Gawinowski; Camille Anoraud
Publié dans: Journal of Air Transport Management, Numéro October 2022, 2022, ISSN 1873-2089
Éditeur: Elsevier
DOI: 10.1016/j.jairtraman.2022.102284

Livrables

Publications

Partager cette page

Télécharger