Towards uncertainty-quantified data-driven modelling and robust control of a wave energy converter prototype

• Data-Driven Modelling - Develops a model for a 1:20-scale Wavestar-type WEC prototype using experimental data collected at Aalborg University’s wave basin. • Uncertainty Quantification - Introduces a polytopic parameterisation to capture and incorporate model uncertainty into the control framework. • Robust Control Synthesis - Formulates a causal H 2 -norm-based controller through linear matrix inequalities derived via Lyapunov theory and Finsler’s Lemma. • Performance Improvement - Demonstrates higher total energy production compared with passive, reactive and H ∞ benchmark controllers. This paper presents a novel unconstrained, causal, and robust H 2 -norm-based control framework for wave energy converters. A data-driven model obtained from experimental measurements on a 1:20-scale Wavestar-type prototype is employed, and model uncertainty is captured through a polytopic parametrisation that defines the structure of the controlled system. Based on Lyapunov theory and Finsler’s Lemma, the control synthesis conditions are formulated as linear matrix inequalities evaluated at the vertices of the polytope. The proposed controller aims to maximise energy extraction by capturing the average closed-loop energy response, providing an effective strategy for control under stochastic, broadband wave conditions. Numerical experiments demonstrate that the controller achieves efficient energy capture without violating physical constraints and outperforms the reactive, passive, and H ∞ benchmark controllers.