A data-driven model-free adaptive controller with application to wind turbines

M L Corradini; G Ippoliti; G Orlando; D Corona

doi:10.1016/j.isatra.2022.11.002

A data-driven model-free adaptive controller with application to wind turbines

ISA Trans. 2023 May:136:267-274. doi: 10.1016/j.isatra.2022.11.002. Epub 2022 Nov 10.

Authors

M L Corradini¹, G Ippoliti², G Orlando³, D Corona⁴

Affiliations

¹ Università di Camerino, 62032 Camerino (MC), Italy. Electronic address: letizia.corradini@unicam.it.
² Università Politecnica delle Marche, 60131 Ancona, Italy. Electronic address: gianluca.ippoliti@univpm.it.
³ Università Politecnica delle Marche, 60131 Ancona, Italy. Electronic address: giuseppe.orlando@univpm.it.
⁴ Università di Camerino, 62032 Camerino (MC), Italy. Electronic address: dario.corona@unicam.it.

PMID: 36437173
DOI: 10.1016/j.isatra.2022.11.002

Abstract

A data-driven controller is presented in this paper, which stems from the well known model-free adaptive control approach based on an equivalent linearized dynamical model of the plant. Inspired by the recent paper (Liu and Yang, 2019), the output tracking problem is here solved by a data-driven adaptive sliding-mode controller simultaneously ensuring prescribed performance constraints. To allow a rigorous stability analysis, the sliding variable, and the consequently derived controller, have been redesigned with respect to the inspiring paper. A proper setting of the gain of the discontinuous term is shown necessary to ensure closed loop stability. Validation of the technique has been extensively performed on the well assessed high-fidelity tool FAST (NREL) to solve the efficiency maximization problem using the proposed approach for a 5 MW wind turbine operating in the medium wind speed region.

Keywords: Data-driven based control; Discrete time sliding mode control; Model-free adaptive control; Prescribed performance control.