Identification and Model Predictive Control of an experimental adaptive optics setup utilizing Kautz basis functions

In this paper, we develop an identification technique based on continuous-time Kautz basis functions and Maximum Likelihood estimation from discrete-time data to obtain a continuous-time model of a laboratory adaptive optics system. We illustrate the proposed identification method using synthetic data and experimental data of a laboratory adaptive optics setup. Finally we utilize the estimated model to develop a Model Predictive Control strategy that considers the deformable mirror actuation constraints. We illustrate the benefits of the model predictive control strategy via simulations and compare it against the classical Proportional-Integral controller.