Data-driven adaptive optimal output-feedback control of a 2-DOF helicopter

This paper studies a data-driven adaptive optimal control problem of a Quanser's 2-degree-of-freedom (DOF) helicopter via output-feedback. A novel sampled-data-based approximate/adaptive dynamic programming (ADP) approach is developed. We start from a stabilizing controller computed using the bound of model uncertainties. Then the optimal control gain is iteratively learned by input/output information. The convergence of the proposed approach is theoretically ensured and the tradeoff between optimality and sampling period is rigorously studied as well. Finally, we show the performance of the proposed algorithm under bounded model uncertainties.