A machine learning approach on past ADS-B data to predict planning controller’s actions
En-route airspace is one of the most congested airspaces, as it is mainly used in the cruise phase of the flight. The en-route sector is usually managed by a team of two air traffic controllers: planning controller (D-side) and executive controller (R-side). The D-side controller is responsible for...
Saved in:
Main Authors: | , , , |
---|---|
Other Authors: | |
Format: | Conference or Workshop Item |
Language: | English |
Published: |
2021
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/146669 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
id |
sg-ntu-dr.10356-146669 |
---|---|
record_format |
dspace |
spelling |
sg-ntu-dr.10356-1466692021-03-12T02:53:05Z A machine learning approach on past ADS-B data to predict planning controller’s actions Pham, Duc-Thinh Alam, Sameer Su, Yi-Lin Duong, Vu N. School of Mechanical and Aerospace Engineering 8th International Conference on Research in Air Transportation (ICRAT ’18) Air Traffic Management Research Institute Engineering::Aeronautical engineering::Air navigation En-route airspace is one of the most congested airspaces, as it is mainly used in the cruise phase of the flight. The en-route sector is usually managed by a team of two air traffic controllers: planning controller (D-side) and executive controller (R-side). The D-side controller is responsible for processing flight plan information to plan and organize the flow of traffic entering the sector. R-side controller deals with ensuring the safety of flights in their sector. A better understanding and predictability of Dside controller actions, for a given traffic scenario, may help in automating some of its tasks and hence reduce workload. In this paper, we propose a learning model to predict D-side controller actions. The learning problem is modeled as a supervised learning problem where the target variables are D-side controller actions and the explanatory variables are the aircraft 4D trajectory features. The model is trained on one month of ADS-B data over an en-route sector, and its generalization performance was assessed, using cross-fold validation, in the same sectors. Results indicate that the model for vertical maneuver actions provides the highest prediction accuracy (99.7%). Besides, the model for speed change and heading change action provides predictability accuracy of 88.7% and 72.4% respectively. The model to predict the set of all the actions (altitude, speed, and heading change) for each flight achieves an accuracy of 0.68 implying for 68% of flights, D-Side Controller’s can be predicted for all the actions from trajectory information at sector entry position. Civil Aviation Authority of Singapore (CAAS) Accepted version This research has been partially supported under Air Traffic Management Research Institute (NTU-CAAS) Grant No. M4061216.05K 2021-03-04T07:37:44Z 2021-03-04T07:37:44Z 2018 Conference Paper Pham, D., Alam, S., Su, Y. & Duong, V. N. (2018). A machine learning approach on past ADS-B data to predict planning controller’s actions. 8th International Conference on Research in Air Transportation (ICRAT ’18), 80. https://hdl.handle.net/10356/146669 80 en © 2018 ICRAT. All rights reserved. This paper was published in International Conference for Research in Air Transportation (ICRAT) and is made available with permission of ICRAT. application/pdf |
institution |
Nanyang Technological University |
building |
NTU Library |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
NTU Library |
collection |
DR-NTU |
language |
English |
topic |
Engineering::Aeronautical engineering::Air navigation |
spellingShingle |
Engineering::Aeronautical engineering::Air navigation Pham, Duc-Thinh Alam, Sameer Su, Yi-Lin Duong, Vu N. A machine learning approach on past ADS-B data to predict planning controller’s actions |
description |
En-route airspace is one of the most congested airspaces, as it is mainly used in the cruise phase of the flight. The en-route sector is usually managed by a team of two air traffic controllers: planning controller (D-side) and executive controller (R-side). The D-side controller is responsible for processing flight plan information to plan and organize the flow of traffic entering the sector. R-side controller deals with ensuring the safety of flights in their sector. A better understanding and predictability of Dside controller actions, for a given traffic scenario, may help in automating some of its tasks and hence reduce workload. In this paper, we propose a learning model to predict D-side controller actions. The learning problem is modeled as a supervised learning problem where the target variables are D-side controller actions and the explanatory variables are the aircraft 4D trajectory features. The model is trained on one month of ADS-B data over an en-route sector, and its generalization performance was assessed, using cross-fold validation, in the same sectors. Results indicate that the model for vertical maneuver actions provides the highest prediction accuracy (99.7%). Besides, the model for speed change and heading change action provides predictability accuracy of 88.7% and 72.4% respectively. The model to predict the set of all the actions (altitude, speed, and heading change) for each flight achieves an accuracy of 0.68 implying for 68% of flights, D-Side Controller’s can be predicted for all the actions from trajectory information at sector entry position. |
author2 |
School of Mechanical and Aerospace Engineering |
author_facet |
School of Mechanical and Aerospace Engineering Pham, Duc-Thinh Alam, Sameer Su, Yi-Lin Duong, Vu N. |
format |
Conference or Workshop Item |
author |
Pham, Duc-Thinh Alam, Sameer Su, Yi-Lin Duong, Vu N. |
author_sort |
Pham, Duc-Thinh |
title |
A machine learning approach on past ADS-B data to predict planning controller’s actions |
title_short |
A machine learning approach on past ADS-B data to predict planning controller’s actions |
title_full |
A machine learning approach on past ADS-B data to predict planning controller’s actions |
title_fullStr |
A machine learning approach on past ADS-B data to predict planning controller’s actions |
title_full_unstemmed |
A machine learning approach on past ADS-B data to predict planning controller’s actions |
title_sort |
machine learning approach on past ads-b data to predict planning controller’s actions |
publishDate |
2021 |
url |
https://hdl.handle.net/10356/146669 |
_version_ |
1696984364374556672 |