Data-driven sampled-data LQR: Certainty-equivalence control via lifted cost and Riccati analysis

A data-driven design framework for sampled-data LQR is presented, in which the continuous-time plant is regulated by a digital controller with a zero-order hold when its model is unknown. This paper extends Rantzer’s recent work on a data-driven Riccati equation to a sampled-data setting. A data-driven sampled-data Riccati equation is derived, from which its result can be obtained directly from input–state trajectories collected at the sampling instants. The conditions for stability and robustness of the closed-loop system are derived subsequently, adapting the analysis to account for inter-sample behavior via lifted cost formulations. It is demonstrated that, once a sufficiently large set of data trajectories has been collected, the closed-loop system attains stability and robustness comparable to those of the standard sampled-data LQR with full model knowledge.