Autonomous Pop-Up Attack Maneuver Using Imitation Learning

Abstract

This study presents a methodology for developing models that replicate the complex pop-up attack maneuver in air combat, using flight data from a Brazilian Air Force pilot in a 6-degree-of-freedom flight simulator. By applying imitation learning techniques and employing a Long Short-Term Memory (LSTM) network, the research trained models to predict aircraft control inputs through sequences of state-action pairs. The model’s performance, evaluated using Root Mean Squared Error (RMSE) and the coefficient of determination (R²), demonstrated its effectiveness in accurately replicating the maneuver. These findings highlight the potential of deploying such models in fully autonomous aircraft, enhancing autonomous combat systems’ reliability and operational capabilities in real-world scenarios.

@inproceedings{dantas2024autonomous,
  author = {Dantas, Joao P. A.},
  title = {{Autonomous Pop-Up Attack Maneuver Using Imitation Learning}},
  booktitle = {Winter Simulation Conference (WSC), PhD Colloquium},
  year = {2024},
  location = {Orlando, FL, USA}
}