Helicopter Tail Rotor and the Training of a Recurrent Neural Network

This work presents the training of a Recurrent Neural Network (RNN) for the identification of dynamics behaviour in aeronautical systems. The network is used to model flap motion on the tail rotor at determined dynamics conditions. The study the tail rotor performance agrees with the expected outcomes. The modelling is not a straightforward task and the dynamics observed in the rotor display that the model could be a suitable tool for monitoring performance under certain conditions.