An Analysis of Human Interaction and Weather Effects on Aircraft Trajectory Prediction via Artificial Intellegence

Artificial intelligence (AI) is developing strongly and widely applied in many fields, including in the aviation. This work provides a deep analysis of the impact of human interaction (HI) and weather effects (WE) on 4D-trajectory prediction using real data and various AI-based models: linear regression (LR – linear regression), random forest regression (RF – random forest), regression with extreme gradient boosting (XGBoost) and a deep neural network model (DNN-Deep neural network). The datasets used in this work contain ADS-B historical data of flights from Ho Chi Minh to Ha Noi in 14 days (12–26 November 2021) and time-synchronized weather data along the waypoints of each flight. We perform many experiments in different scenarios to evaluate the learning performance of LR, RF, XGBoost and DNN models against human interaction (HI) and weather effects (WE). The results prove that DNN model gives the highest accuracy and learning speed among AI-based models, but it is most sensitive to strange information of expert actions and/or weather changes.