Shape Optimization of a Point Absorber Wave Energy Converter for Reduced Current Drag and Improved Wave Energy Capture Using Neural Networks and Genetic Algorithms

Abstract

The shape of the floating buoy of a point absorber wave energy converter (WEC) plays a crucial role in both wave energy harvesting and current drag reduction. In this study, an approach to optimizing the buoy hull geometry with a neural network that replaces the hydrodynamic analysis software is presented, aimed at reducing the ocean current drag force while improving wave energy captured. A new parametric model is introduced to describe the complex shape of the buoy by utilizing the control points of non-uniform rational b-splines. A neural network is developed to significantly reduce the computational time compared to traditional hydrodynamic simulation methods. The optimal hull shape of the buoy is determined by solving an optimization problem using a genetic algorithm, a global optimization technique. The results of the case studies show that the optimal buoy hull shape reduces 68.7% and 71.1% of the current drag, and 50% of mooring line forces compared to the cylinder-shaped buoy and the optimal-power-shaped hull from literature. The optimal buoy hull shape increases the wave energy extraction by 46.1% compared to the thin-ship-shaped buoy but performs 21.1% worse than the optimal-power-shaped hull from the literature.