Implementation of machine learning techniques for the analysis of wave energy conversion systems: a comprehensive review

Abstract

In recent years, marine energy research, like many other branches of science and engineering, has explored the use of increasingly advanced machine learning techniques. Data-driven and machine learning techniques have been shown to be particularly useful in investigating the complex fluid–structure interactions between electromechanical and hydraulic systems and ocean waves. This work provides a comprehensive review of studies that have implemented machine learning and data-driven approaches for system modeling, developing control algorithms, optimizing the system using data-driven modeling, forecasting power generation, and conducting modeling and optimization of arrays of wave energy converters (WECs). The paper briefly discusses various wave energy conversion approaches along with the machine learning techniques typically used in wave energy research. The literature is divided into three main areas: WEC modeling, modeling of WEC arrays, and works focused on forecasting wave characteristics to evaluate the performance of WECs. Finally, the paper discusses the prospective research and development of data-driven and machine learning techniques in this field. The review encompasses literature published between 2008 and 2022.