Data-Driven Model Discovery and Control: Real-Time Implementation to Highly Maneuverable Aircraft Lateral-Directional Dynamics

The aim of this study is to utilize the Dynamic Mode Decomposition with Control (DMDc), a data-driven model discov-ery technique, to obtain a model for the lateral-directional motion of a highly maneuverable aircraft for the purpose of imple-menting integral action based Model Predictive Control (MPC). The model's ability to withstand noisy conditions is evaluated, and its practicality is assessed by analyzing its computational requirements. The computational time and storage requirements are assessed under minimal system requirements. Real-time testing on Speedgoat equipment indicates that the DMDc-MPC algorithm has low computational costs and is suitable for systems with limited computational power and portability.