Abnormal Behavior Detection For Aircraft In Maneuvering Area Based On Surveillance Data

With the number of aircraft take-offs and landings increasing year by year,the airport expansion makes the structure of the airport maneuvering area complex.It also results in the increasing workload of controllers,and the security risk of the airport maneuvering area is becoming higher and higher.Based on the historical data of ADS-B,this paper studies the motion state of the maneuvering area aircraft and detects the abnormal behavior of the maneuvering area aircraft.It can improve the efficiency of the safe of the maneuvering area,reduce the workload of the controller and provide intelligent decision-making for controller.Firstly,the characteristics of aircraft operation under surveillance data were analyzed,and the definition and classification of abnormal behavior of aircraft were carried out in combination with the research results in other fields.Furthermore,abnormal behavior of aircraft in maneuvering area was defined and classified.Secondly,the flight selection,deletion,longitude and latitude conversion,linear difference of missing data were carried out,and continuous track data were obtained.By determining the taxiing paths of the aircraft in the maneuvering area to distinguish the incoming and departure flights.It provided data support for studying abnormal behavior of aircraft in maneuvering area.Finally,the abnormal behavior of the aircraft was studied from two aspects: the speed and the trajectory of the aircraft in the maneuvering area.In the aspect of velocity anomaly detection,the velocity anomaly was classified,the method of determining velocity anomaly was determined,and the model of aircraft velocity anomaly detection in maneuvering area was established,which was considered from the two angles of aircraft glide position and aircraft type.Support vector machine(SVM)was used to detect aircraft velocity anomalies in maneuvering area.In the aspect of trajectory anomaly detection,the mainstream taxiing path of aircraft in maneuvering area was determined by statistics of historical data,and the abnormal trajectory detection model of mainstream taxiway was established.The clustering method based on k-means was used to detect the abnormal trajectory.The experimental results show that this method could quickly detect the abnormal trajectory in the mainstream glide path.