Study on the Identification of Terminal Area Traffic Congestion Situation Based on Symmetrical Random Forest
Abstract
:1. Introduction
2. Modeling of Air Traffic Congestion Situation in the Terminal Area Based on Complex Networks
2.1. Analysis of the Traffic Congestion Situation in the Terminal Area
- First, it is complex. There are a large number of arrival and departure points, waypoints, flight paths, etc., in a large terminal area, and the flight path planning and intersection settings for aircraft flying in and out of different airports are also complicated.
- Second, it is dynamic in nature. The adjustment of aircraft speed during operation, changes in flight altitude, path deviations, and the complex spatiotemporal interdependence between different flight segments cause the congestion situation to be highly dynamic over a short period of time.
- Third, it is global in nature. The terminal area is a large airspace set up to guide multiple aircraft in and out of different airports in an orderly manner. The mutual influence of each aircraft constitutes the congestion situation in the entire terminal area. Therefore, when analyzing it, we must not only focus on the microscopic perspective but also grasp it from a global perspective.
2.2. Definition of Nodes and Edges
2.2.1. Physical Significance
2.2.2. Edge-Weight Setting
- Extraction of busyness (edge weight) recognition indicators;
- Busyness (edge weight) recognition;
3. Extraction of Indicators for Recognizing Traffic Congestion Situations in the Terminal Area
- Average path length ;
- Average point strength ;
- Degree distribution ;
- Network density ;
- Network efficiency ;
4. Terminal Area Traffic Congestion Situation Recognition Method
4.1. Multiclass Random Forest Algorithm
- Construct the feature matrix and the label vector that characterize the traffic congestion situation in the terminal area:
- Construct a single decision tree model;
- Construct a random forest;
4.2. Congestion Situation Recognition Process
- Data collection and collation: collect and collate data on the position of each node in the terminal area and the connecting edges, as well as the position, altitude, heading, etc., of aircraft over a period of time.
- Calculate the busyness of the connecting edges (edge weights). Calculate the busyness of each connecting edge based on the aircraft location information at different times and convert the busyness into edge weights.
- Calculate the terminal area congestion situation recognition index. Calculate the overall terminal area congestion situation recognition index for each period based on the edge weights obtained in step 2.
- Recognize the overall terminal area traffic congestion situation. Use the random forest algorithm to recognize the overall terminal area traffic congestion situation and compare it with the actual congestion situation assessed by experts.
5. Simulation Analysis
5.1. Data Collection and Calculation of Busyness (Side Weight)
5.2. Training of a Congestion Situation Recognition Model
5.3. Congestion Situation Identification Result Analysis
6. Conclusions
- Using a complex network to abstract the elements of the terminal area into nodes and edges, and using the degree of busyness as the edge weight, can closely match the actual situation in the terminal area and lay the foundation for subsequent congestion situation recognition.
- A multiclass random forest algorithm is proposed which improves the performance of a single decision tree compared to the traditional random forest algorithm, and, in turn, improves the accuracy of congestion situation recognition, laying the groundwork for subsequent aircraft deployment and congestion mitigation.
- Visual elements were added to the congestion situation recognition process to facilitate controllers’ identification of busy flight segments and, thus, the rapid formulation of flow control plans.
- The model uses the actual terminal area airspace structure and measured aircraft operation data to train the congestion situation recognition model, which can objectively and truly reflect actual operation.
Author Contributions
Funding
Data Availability Statement
Conflicts of Interest
References
- Xu, X.; Ren, J.; Li, N. Traffic Situation Recognition in Terminal Area Based on FCM. Aeronaut. Comput. Technol. 2014, 44, 1–4. [Google Scholar]
- Zhao, Y.; Lv, L.; Zhang, X. Airspace sector operation status assessment. Sci. Technol. Eng. 2014, 14, 105–109. [Google Scholar]
- Yuan, L.; Hu, M.; Zhang, H.; Ma, Y. State recognition of traffic flow in the Terminal Area based on the fusion of prior experience clustering. J. Transp. Eng. 2016, 16, 83–94. [Google Scholar]
- Wu, M.; Wang, Z.; Gan, X.; Yang, G.; Wen, X. Identification of critical flight conflict points based on complex network theory. J. Northwest. Polytech. Univ. 2020, 38, 279–287. [Google Scholar] [CrossRef]
- Zhao, Y.; Chen, K. Northeastern Univ, Air Traffic Congestion Assessment Method Based on Evidence Theory. In Proceedings of the 2010 Chinese Control and Decision Conference, Xuzhou, China, 26–28 May 2010; Volume 1–5, pp. 426–429. [Google Scholar]
- Liu, F.; Li, J.; Wen, X.; Wang, Y.; Tong, R.; Liu, S.; Chen, D. Situation Assessment of Air Traffic Based on Complex Network Theory and Ensemble Learning. Appl. Sci. 2023, 13, 11957. [Google Scholar] [CrossRef]
- Ding, S.; Huang, H.; Yu, J.; Zhao, H. Research on the hybrid models of granular computing and support vector machine. Artif. Intell. Rev. 2015, 43, 565–577. [Google Scholar] [CrossRef]
- Zhu, C.; Zhang, C.; Guan, J. Airspace traffic situation recognition method based on improved support vector machine. Traffic Inf. Saf. 2023, 41, 76–85. [Google Scholar]
- Yuan, Y.; Chen, B.; Xu, T.; Wang, J. A multi-class CVM flight delay early warning model based on prior knowledge. J. Jilin Univ. Eng. Ed. 2010, 40, 752–757. [Google Scholar]
- Yuan, Y. Research and Application of the Minimum Closed Ball SVM Algorithm. Master’s Thesis, Nanjing University of Aeronautics and Astronautics, Nanjing, China, 2011. [Google Scholar]
- Li, G. Research on Traffic Operation Situation Recognition and Prediction Technology in Airway Networks Based on Track Data. Ph.D. Thesis, Nanjing University of Aeronautics and Astronautics, Nanjing, China, 2020. [Google Scholar]
- Li, Y. Research on Key Technologies for Airspace Resource Allocation in Complex Airport Terminal Areas. Ph.D. Thesis, Nanjing University of Aeronautics and Astronautics, Nanjing, China, 2016. [Google Scholar]
- Jiang, J.; Zhang, H.; Qiu, Q. Evaluation method of air traffic congestion status for traffic flow entering the Terminal Area. J. Wuhan Univ. Technol. Transp. Sci. Eng. Ed. 2016, 40, 1043–1049. [Google Scholar]
- Jiang, J. Research on the Evolution Mechanism of Air Traffic Congestion in the Terminal Area. Master’s Thesis, Nanjing University of Aeronautics and Astronautics, Nanjing, China, 2017. [Google Scholar]
- Wang, H.; Ma, L. Complexity assessment and evolution analysis of traffic in Terminal Areas. Sci. Technol. Eng. 2024, 24, 9235–9243. [Google Scholar]
- Tang, K.; Xu, T.; Guo, T. Research on the topological characteristics and resilience of China’s high-speed railway network from a hypernetwork perspective. Syst. Eng. Theory Pract. 2024, 44, 3639–3649. [Google Scholar]
- Peng, Y.; Wen, X.; Wu, M.; Yang, Z.; Wang, N. Complexity analysis of air traffic based on complex networks in TBO mode. J. Beijing Univ. Aeronaut. Astronaut. 2023, 1–17. [Google Scholar] [CrossRef]
- Peng, Y.; Wen, X.; Wu, M.; Zhu, D.; Ding, L. Flight situation assessment based on complex networks and MFIM-TOPSIS. J. Northwest. Polytech. Univ. 2024, 42, 435–445. [Google Scholar] [CrossRef]
- Breiman, L. Random forests. Mach. Learn. 2001, 45, 5–32. [Google Scholar] [CrossRef]
Period | Edge | (×1000) | Weight | ||
---|---|---|---|---|---|
13:00–13:10 | AAout23-1 | 0.02603 | 0.12608 | 0.07606 | 25% |
ADTout1-5 | 0.39810 | 0.62545 | 0.51178 | 100% | |
ADTout1-4 | 0.00965 | 0.28622 | 0.14794 | 50% | |
AAout4-4 | 0.02781 | 0.00000 | 0.01391 | 25% | |
AAout22-1 | 0.00994 | 0.04463 | 0.02729 | 25% | |
ADout4-2 | 0.11040 | 0.43870 | 0.27455 | 75% | |
TJout6-3 | 0.00251 | 0.06443 | 0.03347 | 25% | |
AADout18-1 | 0.02180 | 0.09422 | 0.05801 | 25% | |
AAout13-1 | 0.17930 | 0.00000 | 0.08965 | 25% | |
ADout12-1 | 0.01829 | 0.23617 | 0.12723 | 50% |
Period | L(in) | L(out) | V | P(in) | P(out) | D | E(in) | E(out) |
---|---|---|---|---|---|---|---|---|
08:00–08:10 | 6.27771 | 15.87777 | 0.21857 | 0.07711 | 0.51618 | 0.00031 | 2.17624 | 6.66134 |
13:00–13:10 | 8.33333 | 12.00000 | 0.17680 | 0.09669 | 0.10221 | 0.00024 | 1.38100 | 2.35790 |
16:00–16:10 | 12.24733 | 13.26661 | 0.30618 | 0.13570 | 0.09821 | 0.00047 | 4.86137 | 5.29014 |
20:00–20:10 | 13.26612 | 8.91934 | 0.14893 | 0.07971 | 0.13713 | 0.00021 | 2.09713 | 2.11371 |
24:00–24:10 | 14.64031 | 8.00136 | 0.24713 | 0.41253 | 0.08126 | 0.00028 | 6.78140 | 1.89714 |
Sample | L(in) | L(out) | V | P(in) | P(out) | D | E(in) | E(out) | Situation |
---|---|---|---|---|---|---|---|---|---|
Sample a | 6.27771 | 15.87777 | 0.21857 | 0.07711 | 0.51618 | 0.00031 | 2.17624 | 6.66134 | |
Sample b | 8.33333 | 12.00000 | 0.17680 | 0.09669 | 0.10221 | 0.00024 | 1.38100 | 2.35790 | |
Sample c | 12.24733 | 13.26661 | 0.30618 | 0.13570 | 0.09821 | 0.00047 | 4.86137 | 5.29014 | |
Sample d | 13.26612 | 8.91934 | 0.14893 | 0.07971 | 0.13713 | 0.00021 | 2.09713 | 2.11371 | |
Sample e | 14.64031 | 8.00136 | 0.24713 | 0.41253 | 0.08126 | 0.00028 | 6.78140 | 1.89714 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Ji, Y.; Yu, F.; Shen, D.; Peng, Y. Study on the Identification of Terminal Area Traffic Congestion Situation Based on Symmetrical Random Forest. Symmetry 2025, 17, 96. https://doi.org/10.3390/sym17010096
Ji Y, Yu F, Shen D, Peng Y. Study on the Identification of Terminal Area Traffic Congestion Situation Based on Symmetrical Random Forest. Symmetry. 2025; 17(1):96. https://doi.org/10.3390/sym17010096
Chicago/Turabian StyleJi, Yuren, Fuping Yu, Di Shen, and Yating Peng. 2025. "Study on the Identification of Terminal Area Traffic Congestion Situation Based on Symmetrical Random Forest" Symmetry 17, no. 1: 96. https://doi.org/10.3390/sym17010096
APA StyleJi, Y., Yu, F., Shen, D., & Peng, Y. (2025). Study on the Identification of Terminal Area Traffic Congestion Situation Based on Symmetrical Random Forest. Symmetry, 17(1), 96. https://doi.org/10.3390/sym17010096