Please use this identifier to cite or link to this item:
Title: A machine learning framework for predicting ATC conflict resolution strategies for conformal automation
Authors: Guleria, Yash
Tran, Phu
Pham, Duc-Thinh
Durand, Nicolas
Alam, Sameer
Keywords: Engineering::Computer science and engineering
Issue Date: 2021
Source: Guleria, Y., Tran, P., Pham, D., Durand, N. & Alam, S. (2021). A machine learning framework for predicting ATC conflict resolution strategies for conformal automation. 11th SESAR Innovation Days (SIDs 2021), 2021-85.
Abstract: Conformal automation allows for increased acceptability of automation tools in air traffic control. The key enabler for achieving conformity of automation tools in performing expert tasks, for example, air traffic conflict resolution, is the identification of ATCO preferences (conflict resolution strategies) and its ability to learn and recommend similar strategies. This research proposes a machine learning-based framework to learn and predict the air traffic conflict resolution strategies using an ensemble model of regressor and classifier chains. This framework enables the prediction and generation of a complete conflict resolution profile of the maneuvered aircraft. Similar and contrasting ATCO conflict resolution strategies are collected through human-in-the-loop experiments, using a real-time, high fidelity simulation environment, for model training and evaluation. The prediction results demonstrate that the ATCOs strategies encoded in the collected data can be learned by the model with high accuracy (95.1%, 93.7% for choice of aircraft) and low MAE( 0.38 Nm and 0.52 Nm for maneuver initiation distance) for the ATCOs' datasets. These results demonstrate high conformance of the model predicted maneuvered trajectories with the original ATCOs maneuvers.
Rights: © 2021 SESAR 3 Joint Undertaking. All rights reserved. This paper was published in Proceedings of 11th SESAR Innovation Days (SIDs 2021) and is made available with permission of SESAR 3 Joint Undertaking.
Fulltext Permission: open
Fulltext Availability: With Fulltext
Appears in Collections:ATMRI Conference Papers
MAE Conference Papers

Files in This Item:
File Description SizeFormat 
A_Machine_Learning_Framework_for_Identifying_and_Predicting_ATC_Conflict_Resolution_Strategies.pdfMain article3.23 MBAdobe PDFThumbnail

Page view(s)

Updated on May 19, 2022

Download(s) 50

Updated on May 19, 2022

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.