Benchmarking control strategies for wave energy converters: From idealised knowledge to realistic operation

Abstract

Maximising energy extraction in wave energy converters remains challenging due to the non-causal nature of the optimal control, complex interactions and real-time constraints. This paper presents a benchmarking of four strategies: LiTe-Con, LiTe-Con+, model predictive control (MPC) and spectral control. The evaluation transitions from idealised scenarios with perfect knowledge of the wave excitation force to realistic conditions accounting for estimation and forecasting. The study is conducted using a 1:20 scale pitching WEC model subjected to representative wave climates in the Gulf of Biscay. Results quantify performance degradation induced by imperfect information, highlighting trade-offs between absorbed energy, robustness, and computational complexity. While optimisation-based controllers (MPC and spectral) provide systematic constraint handling and high performance under ideal conditions, they are sensitive to prediction accuracy, with performance reducing 3.27% and 5.1%, respectively, under realistic conditions. In contrast, LiTe-Con+ achieves comparable energy absorption (only 1.57% reduction) with significantly lower computational requirements (being 25 and 930 times faster than spectral and MPC, respectively) and improved robustness, making it a strong candidate for practical controller selection in real WEC deployments.