1: \begin{abstract}

2: This paper addresses a data-driven input reconstruction problem based on Willems’ Fundamental Lemma in which unknown input estimators (UIEs) are constructed directly from historical I/O data. Given only output measurements, the inputs are estimated by the UIE, which is shown to asymptotically converge to the true input without knowing the initial conditions. Both open-loop and closed-loop UIEs are developed based on Lyapunov conditions and the Luenberger-observer-type feedback, whose convergence properties are studied. An experimental study is presented demonstrating the efficacy of the closed-loop UIE for estimating the occupancy of a building on the EPFL campus via measured carbon dioxide levels.

3: 

4: % This paper addresses a data-driven input reconstruction problem based on the Willems' Fundamental Lemma. Two different approaches are proposed to enable the direct design of an unknown input estimator from historical I/O data. These two methods are respectively developed based on the Lyapunov condition and the Luenberger-observer-type feedback, whose convergence properties are studied. The proposed approaches are validated by simulations and by an experiment in a real-world building, which estimates the number of occupants by the indoor CO2 level.

5: 

6: % Given the output measurement, the estimated input by the UIE asymptotically converges to the actual input without knowing the initial conditions. Different open-loop and closed-loop UIEs are elaborated based on the Lyapunov condition and generalized inverse of Hankel matrices, while their existence conditions are provided. Apart from a simulation validation, we apply the closed-loop UIE to estimate the occupant number by carbon dioxide levels in an actual building.

7: 

8: \end{abstract}

9: