Standard Procedure-Guided Flight Trajectory Pattern Mining for Airport Terminal Airspace

Terminal airspace complexity necessitates adherence to standard procedures and instrument flight rules for aircraft arrivals and departures. Standard procedures are responsible for defining the planned trajectories for aircraft under normal conditions. Therefore, this study proposes a method for mining flight trajectory patterns in airport terminal airspace based on standard procedure guidance. Dynamic time warping measures trajectory similarity to address temporal mismatches between standard flight procedures and actual trajectories. This method identifies complete matching trajectories that describe standard flight procedures and facilitates subsequent clustering. In the adaptive K-Medoid clustering process, cluster centers are initialized and updated based on the guidance provided by matching trajectories from standard procedures. We compare the proposed method with traditional clustering methods using Lukou Airport in Nanjing, China as a case study. Guided by standard flight procedures, the proposed method achieves a silhouette coefficient of 0.678 for arrival flight trajectory patterns and 0.661 for departure flight trajectory patterns, as indicated by the results. Moreover, the CH, SP, and CP indexes outperform the comparative experimental methods, demonstrating the improved clustering effectiveness. Additionally, by comparing the results on a subset of data that is difficult to cluster using traditional methods, the advantages of the proposed method in handling such datasets are proven. The proposed method aligns better with the current airspace conditions, improving the efficiency and rationality of mining flight trajectory patterns in terminal airspace.