Machine learning evaluation of wind, solar, and hybrid off-grid residential energy systems under matched renewable generation


AYAZ ATALAN Y., ATALAN A.

Sustainable Energy Technologies and Assessments, cilt.95, 2026 (SCI-Expanded, Scopus)

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 95
  • Basım Tarihi: 2026
  • Doi Numarası: 10.1016/j.seta.2026.105567
  • Dergi Adı: Sustainable Energy Technologies and Assessments
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Compendex, Geobase, INSPEC
  • Anahtar Kelimeler: Energy management system, Extra trees regression, Hybrid renewable energy, Machine learning, Off-grid energy systems
  • Çanakkale Onsekiz Mart Üniversitesi Adresli: Evet

Özet

Off-grid residential system design often prioritizes annual energy sufficiency while neglecting temporal resource complementarity, creating a critical research gap in reliability assessment under strictly matched generation constraints. This study evaluates wind-only, photovoltaic (PV), and hybrid wind/PV architectures for an off-grid residence using an identical annual renewable energy budget and a dynamic hourly load profile. The primary novelty lies in isolating the effect of temporal complementarity by enforcing a matched-energy constraint across all architectures, combined with a target-specific machine learning (ML) forecasting interface for the energy management system (EMS). The framework utilized real geographic coordinates of Çanakkale, Türkiye, to generate a physically constrained 8,760-hour simulation. Five ML methods evaluate 24-hour prior predictions for PV, wind, and load. Quantitatively, the hybrid system meets 99.065% of annual demand, with a loss-of-power-supply probability (LPSP) of 0.935%, significantly outperforming wind-only (82.948% served, 17.052% LPSP) and PV-only (84.0% served, 16.0% LPSP) systems. Extra Trees achieves test R2 values of 0.989 for PV and 0.841 for wind, while XGBoost reaches 0.624 for the load. Qualitatively, these findings demonstrate that complementary generation timing substantially reduces temporal mismatch and curtailment of energy without increasing the renewable energy budget. House-scale off-grid design must jointly evaluate source complementarity, dynamic demand, and forecast uncertainty to ensure reliable operation.