Multi-objective optimization of co-located wave-wind farm layouts supported by genetic algorithms and numerical models

• Published in: Renewable energy 2025-03, Vol.241, p.122362, Article 122362
• Main Authors: Teixeira-Duarte, Felipe, Rosa-Santos, Paulo, Taveira-Pinto, Francisco
• Format: Article
• Language: English
• Subjects: Co-located wave-wind energy farms; Genetic algorithm; K-means; Layout optimization; Offshore renewable energy; SNL-SWAN; WEC
• ISSN: 0960-1481
• DOI: 10.1016/j.renene.2025.122362

This study introduces a novel methodology for optimizing Wave Energy Converter (WEC) positioning in an array using a continuous domain, surpassing the traditional fixed layout approaches. The Wave Energy Park Layout Assessment Index (WLA), which integrates the wave protection factor (HRA) and power absorption efficiency (q-factor), is employed to evaluate the performance of WEC farms. To enhance computational efficiency, unsupervised classification methods, such as k-means clustering, are used to reduce the number of sea states while accurately representing wave energy, preserving 90 % of incoming wave energy. Genetic algorithms, integrating the SNL-SWAN hydrodynamic model, are then used to optimize WEC layout by balancing exploration and computational cost, maintaining solution diversity, and avoiding premature convergence. Compared to the non-optimized designs, the proposed method increases absorbed wave power by 87 % and wave height reduction by 46 %. The study acknowledges trade-offs between objectives and area restrictions, and provides an open-source code for further research and development in WEC farm optimization. This integrated approach aims to enhance the efficiency and effectiveness of WEC farm designs, offering a robust framework for future advancements in wave energy extraction. [Display omitted] •Genetic algorithms used to optimize wave farm layouts by balancing exploration and computational cost.•A flexible continuous domain used for WEC positioning, improving upon previous grid-based layout methodologies.•K-means clustering effectively increases computational efficiency while preserving 90 % of incoming wave energy.•Increase in absorbed wave power (87 %) and wave height reduction (46 %) compared to previous studies without optimization.•Open-source code and processing tools made available on GitHub to accelerate development of efficient wave farms.