Trajectory clustering method research and application

In order to improve terminal arrival route, increase terminal area capacity, enhance security level of air traffic management and air traffic controller´s operating efficiency. Actual trajectory data were clustered based on kmeans algorithm. Trajectory data is assigned to several different clusters. Mean arrival routes can be obtained from trajectory clustering and there are deviation between mean arrival routes and Standard Terminal Arrival Route (STAR). Analysis of the main factors of the deviation and proposes three evaluation parameters for the deviation. This paper provides some references for the redesign of flight procedures.