Analysis of flight operation characteristics based on DBSCAN and K-Means

With the increasing complexity of air traffic, the operational characteristics of flights remain largely unexplored. In particular, the revision of Scheduled Flight Block Time (SFBT) heavily relies on statistical analysis of historical data. Therefore, the objective of this paper is to propose a method for analyzing flight operation characteristics from a spatial-temporal perspective. To achieve this, the DBSCAN algorithm was employed to uncover spatial aggregation patterns among flight segments. Additionally, the K-Means algorithm was utilized to investigate the periodicity of flight block time. Based on our findings, it is observed that the majority of airport segments can be categorized into 4-5 distinct groups. Furthermore, it was discovered that taxi time exhibits a higher degree of periodicity compared to flight air time. Overall, these results provide valuable insights into the characteristics of flight operations, shedding light on the overlooked aspects of air traffic management.