AI-driven Predictive Load Management in Hybrid Microgrids in the Reduction of Dependence Towards Diesel Generators on Remote Indonesian Islands (3T)
Year: 2026
Type: article
Abstract: This literature-based research proposal examines the potential application of Artificial Intelligence-driven Energy Management Systems (AI-EMS) to reduce diesel generator dependency in hybrid microgrids serving Indonesia's Frontier, Outermost, and Underdeveloped (3T) regions. Conventional Rule-Based Energy Management Systems (RB-EMS) face limitations in managing intermittent renewable energy generation, fluctuating electricity demand, and battery storage utilization. To address these challenges, this study proposes a predictive energy management framework integrating AI-based load and solar generation forecasting with Receding Horizon Control (RHC) for optimized energy dispatch. The proposed system coordinates solar photovoltaic generation, Battery Energy Storage Systems (BESS), and diesel generators to improve renewable energy utilization while maintaining microgrid reliability. Drawing upon existing literature and secondary data, the essay examines potential improvements in solar fraction, specific fuel consumption, power supply reliability, and Levelized Cost of Energy (LCOE). It also considers implementation challenges, including computational requirements, data availability, infrastructure limitations, and initial investment costs. This work was developed as an independent research essay for the Research Poster Competition (RPC) 2026. It presents a conceptual engineering framework and literature-based techno-economic assessment rather than original experimental or simulation results. The proposal aims to contribute to discussions on intelligent energy management, sustainable rural electrification, and Indonesia's transition toward lower-diesel energy systems. Disclaimer: This manuscript is an independent, literature-based competition research proposal and has not undergone formal peer review.
Author: Sawarsa, Christopher
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7225429650