A prediction method for bounced landing of aircraft based on DBN

Abstract

In order to grasp causes of bounced landing of aircraft which is a frequently occurring issue during flight operationꎬ and effectively prevent such incidentsꎬ a prediction method for bounced landing based on DBN was proposed. Secondlyꎬ correlation between incidents and landing airports ' environment was evaluated by using aviation dataꎬ and with an actual incident of China Eastern Airlines as an exampleꎬ changing trend of bounced landing along with throttle stick position at touchdown was explored. Thenꎬ impacts of pilot controlꎬ aircraft status and unstable approach on incidents were discussed. Finallyꎬ different combinations of information were used as inputs to train the modelꎬ and their prediction accuracy was compared to find optimal one. The results show that DBN ̄based method is suitable for predicting bounced landing by utilizing flight data. When network input includes direct influencing factors such as airports' environmentꎬ throttle lever positionꎬ as well as indirect ones like unstable approachꎬ this model can accurately predict accidents with a prediction accuracy as high as 94􀆰 78%.