Prediction of Hydrodynamic Coefficients of a Point Absorber Wave Energy Converter Using Artificial Neural Networks and Its Applications

With the growing interest in renewable energy, wave energy has garnered attention for its relatively high energy density and sustainability. Among various types of wave energy converters (WECs), the point-absorber WEC (PT-WEC) stands out as an excellent option for achieving high power generation efficiency due to its ability to generate electricity through the motion of a floating body induced by ocean waves. However, analyzing the power generation of such WECs requires the calculation of the floating body’s motion response (RAO), which is typically performed using Frequency-Domain analysis based on potential flow theory and the Boundary Element Method (FD-BEM). While Frequency-Domain analysis has been extensively studied over decades, it often relies on specialized hydrodynamic software and demands significant computational time and cost. This study aims to develop a computational framework based on Artificial Neural Networks (ANNs) to efficiently and accurately predict the hydrodynamic coefficients of a 2D rectangular floating body. A dataset of hydrodynamic coefficients was generated using an in-house 2D FD-BEM, and training and validation were performed using an ANN linked with the program. The ANN model employs three input variables: the width-to-water depth ratio, the water depth-to-draft ratio, and the dimensionless wavenumber. The output layer provides hydrodynamic coefficients corresponding to the three degrees of freedom motion (surge, heave, and pitch motion). Based on this, a post-processing wrapping program was developed to calculate the RAO and power generation of the WEC. The ANN model and wrapping program were integrated into a unified computational framework, which was applied to predict the power generation of a PT-WEC. The results were compared with those obtained from FD-BEM analysis, demonstrating high accuracy and efficiency. This framework can be extended to account for various floating body geometries and mooring systems in WEC power generation calculations. Furthermore, the methodology can also be applied to analyze the hydrodynamic characteristics of three-dimensional floating bodies, further broadening its potential applications in hydrodynamic analysis and WEC optimization.