Transactions of Nanjing University of Aeronautics & Astronautics
Flight Trajectory Option Set Generation Based on Clustering Algorithms
- Article
- Figures
- Metrics
- Preview PDF
- Reference
- Related
- Cited by
- Materials
Abstract:
Addressing the issue that flight plans between Chinese city pairs typically rely on a single route， lacking alternative paths and posing challenges in responding to emergencies， this study employs the “quantile-inflection point method” to analyze specific deviation trajectories， determine deviation thresholds， and identify commonly used deviation paths. By combining multiple similarity metrics， including Euclidean distance， Hausdorff distance， and sector edit distance， with the density-based spatial clustering of applications with noise （DBSCAN） algorithm， the study clusters deviation trajectories to construct a multi-option trajectory set for city pairs. A case study of 23 578 flight trajectories between the Guangzhou airport cluster and the Shanghai airport cluster demonstrates the effectiveness of the proposed framework. Experimental results show that sector edit distance achieves superior clustering performance compared to Euclidean and Hausdorff distances， with higher silhouette coefficients and lower Davies-Bouldin indices， ensuring better intra-cluster compactness and inter-cluster separation. Based on clustering results， 19 representative trajectory options are identified， covering both nominal and deviation paths， which significantly enhance route diversity and reflect actual flight practices. This provides a practical basis for optimizing flight paths and scheduling， enhancing the flexibility of route selection for flights between city pairs.
Keywords:
Project Supported:
This work was supported in part by Boeing Company and Nanjing University of Aeronautics and Astronautics （NUAA） through the Research on Decision Support Technology of Air Traffic Operation Management in Convective Weather under Project 2022-GT-129， and in part by the Postgraduate Research and Practice Innovation Program of NUAA （No.xcxjh20240709）.
Clc Number:
V355
Articles
Mobile website
