| With the rapid development of China’s civil air transport industry,the flow of flights has continued to grow.The increase of flight traffic in the terminal area has been particularly significant.This has increased the workload of the control unit and may lead to security risks.Therefore,it is necessary to identify the abnormal aircraft behavior in the terminal area.The identification of abnormal aircraft behavior in the terminal area is not only a prerequisite for the development of an efficient decision support system,but it can also provide an effective decision-making scheme for air traffic controller.This paper uses the cluster analysis in data mining technology to complete the aircraft trajectory classification based on the surveillance data in the terminal area.Based on the trajectory classification,a statistical method is used to complete the aircraft abnormal behavior identification in the terminal area.Firstly,introduces the composition and characteristics of the radar data in the terminal area,gives the processing method of the radar data in the terminal area,base on summarizing the trajectory similarity model measurement methods in different fields,and constructs the trajectory similarity based on the speed correction coefficient Model;The trajectory clustering based on the velocity correction coefficient was used to classify the aircraft trajectory in the terminal area.Then,on the basis of cluster analysis,taking into account aircraft trajectory position,speed and other factors,giving the definition of aircraft anomaly in the terminal area and the aircraft energy height anomaly in the terminal area is proposed.The identification method of abnormal aircraft behavior in the terminal area is given.Finally,on the basis of statistical analysis of data structure,by calculating the characteristics of data distribution,the center trajectory and boundary values of the abnormal trajectory identification are obtained,Thus,the abnormal trajectory of the aircraft position and the altitude trajectory of the aircraft energy were identified,and the recognition results were analyzed.Experiments show that the aircraft trajectory spectrum clustering based on the velocity correction coefficient can achieve a reasonable division of aircraft trajectories;and the aircraft abnormal behavior identification based on statistical methods combined with statistical methods can identify aircraft abnormal behavior in the terminal area. |