Iterative feedback tuning and closed-loop output error identification applied to Smith predictors

In this paper, we study the applicability of two algorithms to Smith predictor dead time compensation. The first algorithm is Iterative Feedback Tuning (IFT), a model free gradient descent controller tuning tool. Refer to [1, 2] for further details. The second algorithm is Closed-Loop Output Error (CLOE) identification. This identification algorithm minimizes the difference between the achieved closed-loop response and the response of a closed-loop output error predictor, i.e. it is an identification algorithm that is tailored to closed-loop problems. Refer to [4] for further details.