A machine learning approach for conflict resolution in dense traffic scenarios with uncertainties
With the continuous growth in the air transportation demand, air traffic controllers will have to handle increased traffic and consequently more potential conflicts. This gives rise to the need for conflict resolution advisory tools that can perform well in high-density traffic scenarios given a noi...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Conference Paper |
| Language: | English |
| Published: |
2021
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/146568 |
| _version_ | 1811694850419982336 |
|---|---|
| author | Pham, Duc-Thinh Tran, Ngoc Phu Alam, Sameer Duong, Vu Delahaye, Daniel |
| author2 | School of Mechanical and Aerospace Engineering |
| author_facet | School of Mechanical and Aerospace Engineering Pham, Duc-Thinh Tran, Ngoc Phu Alam, Sameer Duong, Vu Delahaye, Daniel |
| author_sort | Pham, Duc-Thinh |
| collection | NTU |
| description | With the continuous growth in the air transportation demand, air traffic controllers will have to handle increased traffic and consequently more potential conflicts. This gives rise to the need for conflict resolution advisory tools that can perform well in high-density traffic scenarios given a noisy environment. Unlike model-based approaches, learning-based or machine learning approaches can take advantage of historical traffic data and flexibly encapsulate the environmental uncertainty. In this study, we propose an artificial intelligent agent that is capable of resolving conflicts, in the presence of traffic and given uncertainties in conflict resolution maneuvers, without the need for prior knowledge about a set of rules mapping from conflict scenarios to expected actions. The conflict resolution task is formulated as a decision-making problem in a large and complex action space, which is applicable for employing the reinforcement learning algorithm. Our work includes the development of a learning environment, scenario state representation, reward function, and learning algorithm. As a result, the proposed method, inspired by Deep Q-learning and Deep Deterministic Policy Gradient algorithms, can resolve conflicts, with a success rate of over 81%, in the presence of traffic and varying degrees of uncertainties. |
| first_indexed | 2024-10-01T07:14:07Z |
| format | Conference Paper |
| id | ntu-10356/146568 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T07:14:07Z |
| publishDate | 2021 |
| record_format | dspace |
| spelling | ntu-10356/1465682021-03-06T20:10:30Z A machine learning approach for conflict resolution in dense traffic scenarios with uncertainties Pham, Duc-Thinh Tran, Ngoc Phu Alam, Sameer Duong, Vu Delahaye, Daniel School of Mechanical and Aerospace Engineering Thirteenth USA/Europe Air Traffic Management Research and Development Seminar (ATM2019) Air Traffic Management Research Institute Engineering::Aeronautical engineering Air Traffic Command and Control Reinforcement Learning With the continuous growth in the air transportation demand, air traffic controllers will have to handle increased traffic and consequently more potential conflicts. This gives rise to the need for conflict resolution advisory tools that can perform well in high-density traffic scenarios given a noisy environment. Unlike model-based approaches, learning-based or machine learning approaches can take advantage of historical traffic data and flexibly encapsulate the environmental uncertainty. In this study, we propose an artificial intelligent agent that is capable of resolving conflicts, in the presence of traffic and given uncertainties in conflict resolution maneuvers, without the need for prior knowledge about a set of rules mapping from conflict scenarios to expected actions. The conflict resolution task is formulated as a decision-making problem in a large and complex action space, which is applicable for employing the reinforcement learning algorithm. Our work includes the development of a learning environment, scenario state representation, reward function, and learning algorithm. As a result, the proposed method, inspired by Deep Q-learning and Deep Deterministic Policy Gradient algorithms, can resolve conflicts, with a success rate of over 81%, in the presence of traffic and varying degrees of uncertainties. Civil Aviation Authority of Singapore (CAAS) This research is partially supported by Air Traffic Management Research Institute (NTU-CAAS) Grant No. M4062429.052 2021-03-01T07:59:12Z 2021-03-01T07:59:12Z 2019 Conference Paper Pham, D.-T., Tran, N. P., Alam, S., Duong, V., & Delahaye, D. (2019). A machine learning approach for conflict resolution in dense traffic scenarios with uncertainties. Proceedings of Thirteenth USA/Europe Air Traffic Management Research and Development Seminar (ATM2019). https://hdl.handle.net/10356/146568 en © 2021 AI Team. All rights reserved. application/pdf |
| spellingShingle | Engineering::Aeronautical engineering Air Traffic Command and Control Reinforcement Learning Pham, Duc-Thinh Tran, Ngoc Phu Alam, Sameer Duong, Vu Delahaye, Daniel A machine learning approach for conflict resolution in dense traffic scenarios with uncertainties |
| title | A machine learning approach for conflict resolution in dense traffic scenarios with uncertainties |
| title_full | A machine learning approach for conflict resolution in dense traffic scenarios with uncertainties |
| title_fullStr | A machine learning approach for conflict resolution in dense traffic scenarios with uncertainties |
| title_full_unstemmed | A machine learning approach for conflict resolution in dense traffic scenarios with uncertainties |
| title_short | A machine learning approach for conflict resolution in dense traffic scenarios with uncertainties |
| title_sort | machine learning approach for conflict resolution in dense traffic scenarios with uncertainties |
| topic | Engineering::Aeronautical engineering Air Traffic Command and Control Reinforcement Learning |
| url | https://hdl.handle.net/10356/146568 |
| work_keys_str_mv | AT phamducthinh amachinelearningapproachforconflictresolutionindensetrafficscenarioswithuncertainties AT tranngocphu amachinelearningapproachforconflictresolutionindensetrafficscenarioswithuncertainties AT alamsameer amachinelearningapproachforconflictresolutionindensetrafficscenarioswithuncertainties AT duongvu amachinelearningapproachforconflictresolutionindensetrafficscenarioswithuncertainties AT delahayedaniel amachinelearningapproachforconflictresolutionindensetrafficscenarioswithuncertainties AT phamducthinh machinelearningapproachforconflictresolutionindensetrafficscenarioswithuncertainties AT tranngocphu machinelearningapproachforconflictresolutionindensetrafficscenarioswithuncertainties AT alamsameer machinelearningapproachforconflictresolutionindensetrafficscenarioswithuncertainties AT duongvu machinelearningapproachforconflictresolutionindensetrafficscenarioswithuncertainties AT delahayedaniel machinelearningapproachforconflictresolutionindensetrafficscenarioswithuncertainties |