juc JURNAL UJI COBAJUC 1234-273X Karoteh Utama Publisher Intelligent Infrastructure A Hybrid LSTM-Attention Approach to Short-Term Load Forecasting in Island Microgrids ManurungYohanes RahmanAisha Politeknik Digital Borneo Universiti Teknologi Selatan 21092026 1 1 30 42 14052026 23072026 22082026 © 2026 The Authors 2026 Creative Commons Attribution 4.0 International License

Island microgrids run on diesel and solar, and poor demand forecasts waste fuel. We combine an LSTM encoder with temporal attention and weather covariates to forecast load 24 hours ahead for three microgrids. Mean absolute percentage error drops to 4.8%, against 7.9% for SARIMA and 6.1% for a plain LSTM. Simulated dispatch with the new forecasts reduces diesel consumption by 6.4% per year.

 load forecasting LSTM attention microgrid