A Novel Hybrid Optimization Framework for Renewable Energy Investment in Hot Arid Regions: Integrating Time-Series Forecasting and AI-Driven Algorithms for Algeria’s Grid-Export Strategy

##plugins.themes.academic_pro.article.main##

Llahm Omar Ben Dalla
Ömer KARAL
Ali Degirmenci
HÜSEYİN CANBOLAT
Fatih V ÇELEBI
Yasser Fathi Nassar

الملخص

Unlocking the vast solar and wind potential of arid regions like Algeria requires innovative planning paradigms for domestic decarbonization and international exports. This study introduces a novel, comprehensive AI-powered hybrid optimization system integrated within the open-source IRENA FlexTool 3 platform. We developed a unique methodology combining quantized Long Short-Term Memory (LSTM) networks for forecasting (MAE < 6%) with a lightweight transformer-driven stochastic optimizer. This framework co-optimizes grid stability, renewable investment, and HVDC-enabled exports to Europe under extreme uncertainty, utilizing verified high-resolution OASES/LEAP-RE data. The research demonstrates that strategically deploying 12 GW PV and 5 GW wind in the Sahara achieves a levelized export cost of €38/MWh, 18% lower than traditional methods. Furthermore, domestic curtailment decreases from over 25% to 8%, enabling 28 TWh/year of clean exports and displacing 10.6 MtCO₂ annually. This work's novelty lies in its scenario-aware revenue optimization and edge-compatible AI models within a reproducible, open-data environment. It significantly benefits academia and the broader global scientific community by providing a scalable, transparent model for energy system planning. By converting surplus desert generation into affordable green exports, this study advances energy sovereignty and climate action. It offers a robust framework for transcontinental energy justice, establishing a new benchmark for integrating advanced AI into sustainable energy policy and infrastructure development globally. The open-source nature ensures reproducibility, allowing researchers worldwide to adapt this hybrid AI approach for diverse arid contexts, thereby accelerating the transition towards resilient, low-carbon energy systems through data-driven decision-making and fostering collaborative innovation in renewable energy integration strategies.

##plugins.themes.academic_pro.article.details##

كيفية الاقتباس
Dalla, L. O. B., KARAL, Ömer, Degirmenci, A., CANBOLAT, H., ÇELEBI, F. V., & Nassar, Y. F. (2026). A Novel Hybrid Optimization Framework for Renewable Energy Investment in Hot Arid Regions: Integrating Time-Series Forecasting and AI-Driven Algorithms for Algeria’s Grid-Export Strategy. مجلة جامعة فزان العلمية, 62–76. استرجع في من https://fezzanu.edu.ly/fusj/index.php/FUAJ/article/view/784