Mendonça, Maria MendesSantos, Vítor2025-05-162025-05-162025-06-150959-6526PURE: 115423327PURE UUID: 56c98559-6799-44a9-85ca-a9eb03131516Scopus: 105004807991WOS: 001501113400001ORCID: /0000-0002-4223-7079/work/182886984http://hdl.handle.net/10362/183104Mendonça, M. M., & Santos, V. (2025). Advancing sustainable energy solutions: AI hybrid renewable energy systems with hybrid optimization algorithms and multi-objective optimization in Portugal. Journal of Cleaner Production, 511, Article 145564. https://doi.org/10.1016/j.jclepro.2025.145564 --- %ABS1% --- This work has been supported by Portuguese funds through FCT - Fundação para a Ciência e Tecnologia, I.P., under the project FCT UIDB/ 04466/2020, Lisbon, Portugal, and this work has been supported by Information Management Research Center (MagIC) - NOVA Information Management School, Lisbon, Portugal.The effects of global warming are becoming increasingly evident in our daily lives, making it essential to develop sustainable, carbon-neutral solutions. The energy sector is a major contributor to global warming due to its reliance on coal, oil, and natural gas. Therefore, transitioning to renewable energy is crucial. However, renewable sources face intermittency issues, as their availability depends on weather conditions. To address this, a hybrid energy system that integrates multiple renewable sources can enhance reliability. This study optimized a hybrid renewable energy system for Portugal. It began with a systematic literature review on artificial intelligence and energy, identifying nine relevant studies that helped formulate the problem. A hybrid system incorporating solar panels and wind turbines was designed and optimized using a novel algorithm that combined Particle Swarm Optimization (PSO) and Grey Wolf Optimization (GWO). By leveraging the strengths of both methods, the algorithm improved convergence toward the global optimum. To further enhance efficiency, the algorithm was parallelized to reduce execution time and computational demands. Three experiments were conducted to optimize the system. In the first experiment, the Average Best Fitness (ABF) started below 0.420, decreased to 0.390 by the third iteration, but later increased to approximately 0.400. To improve performance, hyperparameters were adjusted in a second experiment. However, results worsened, with ABF starting at 0.480 and only reaching 0.455 by iteration 14. The third experiment yielded the most promising results, with an initial ABF of 0.150, followed by a sharp drop at iteration 3 and a gradual decline with fluctuations. Future research should explore a broader range of hyperparameter combinations to refine optimization results. Additionally, incorporating economic and social objectives alongside technical and environmental criteria will provide a more comprehensive assessment of hybrid energy systems.155232511engHybrid AlgorithmMulti-objective optimizationHybrid Renewable Energy SystemsRenewable Energy, Sustainability and the EnvironmentGeneral Environmental ScienceStrategy and ManagementIndustrial and Manufacturing EngineeringSDG 7 - Affordable and Clean Energyjournal article10.1016/j.jclepro.2025.145564https://www.scopus.com/pages/publications/105004807991https://www.webofscience.com/wos/woscc/full-record/WOS:001501113400001