Faid Rama Daniy, Nicodemus Oloan Simanjuntak, Elya Syafa'atun Ni'mah, Muhammad Nabel Al Fayyed, Rifdah Mayhasna Nur Alayya, Rizky Ajie Aprilianto
The urgent global transition toward sustainable energy requires a reliance on variable renewable sources, particularly solar and wind, whose inherent intermittency poses significant challenges to grid stability and management, necessitating robust forecasting and management. This study conducts a Systematic Literature Review (SLR) of 36 selected research articles from 977 Scopus and ScienceDirect databases to explore the trends, challenges, and effectiveness of Machine Learning (ML) techniques in forecasting renewable energy efficiency. The selection process follows the Population, Intervention, Comparison, Outcome, and Context (PICOC) framework to formulate research questions. Then, it adopts the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology to identify and filter relevant literature. The review highlights a growing trend in applying advanced ML models, such as Random Forest (RF), Support Vector Machine (SVM), Extreme Learning Machine (ELM), and Multilayer Perceptron (MLP), which demonstrate strong predictive capabilities for nonlinear energy patterns. Despite their advantages, several implementation challenges persist, including limited data availability, model interpretability, and computational demands. The findings indicate that ML-based forecasting models can significantly improve prediction accuracy and energy management strategies, offering valuable insights for future research and practical deployment in renewable energy systems. © 2025 IEEE.
Universitas Negeri Semarang, Department of Electrical Engineering, Semarang, Indonesia