Detection of unstable approaches in flight track with recurrent neural network

Due to the increasing of air traffic density is needed to improve flight safety level especially in approach and landing phase. A potentially risky approach phase of flight is called an unstable approach. There are related studies in detecting these conditions, including anomaly detection of flight tracks for all flight phases. Compare with Multiple Kernel and Clustering methods, Recurrent Neural Network (RNN) has advantages in term of accuracy, sensitivity towards short term anomaly, and does not require dimensional reduction to identify the flight track anomaly pattern. However, RNN require experts to identify the risk event characteristics that cause anomalies. While heuristic methods can identify anomaly patterns specifically in the approach phase or called unstable approach with rule-based for each risk event. Therefore, this study combines the two methods approach to identify the unstable approach pattern. The main focus of this research is to prepare the data until ready to do the process of model formation by preprocessing technique, then done data modelling using RNN method with architecture stacked Long Short Term Memory (LSTM), and identify the type of risk event that influence unstable approach with heuristic method. In the modelling experiment performed by tuning the optimum value for each input parameter. The optimum value for batch size = 128, epoch = 150, and optimizer is rmsprop, with accuracy value equal to 90.12%, recall value equal to 59,44%, and precision value equal to 100%. That results show that the established model has been able to classify unstable and stable approaches, with high precision value, but the recall value is still low. Therefore for future work, the more amount of unstable approach data can be used to improve performance of data training.