Robust data-driven control of quadrotors: From theory to practice

This study explores the application of data-driven methods for developing robust flight control systems for unmanned aerial vehicles (UAVs), specifically quadrotors. The theory of data-driven robust control based on the matrix S-lemma and the formulation of the associated linear matrix inequalities (LMIs) are revisited, followed by their application to a six-degrees-of-freedom (6DOF) nonlinear quadrotor system. This work extends previous research on data-driven synthesis of robust control laws using the matrix S-lemma approach through hardware implementation and experimental validation. Firstly, a high-fidelity simulation is employed to verify the derived control law’s ability to stabilize the UAV, delivering closed-loop performance comparable to a model-based linear–quadratic regulator (LQR) design. The hardware implementation and experimental results are presented, further validating the method’s effectiveness in real-world conditions. This research strengthens the connection between theoretical advancements in data-driven control and practical applications in UAV systems, advancing beyond simulation to hardware implementation.