Hybrid solar–battery mini-grids need coordinated decisions on generation, storage, demand and power quality, yet monitoring-only installations cannot decide when to charge the battery, shift flexible loads or respond to a forecast evening peak. This paper develops and evaluates a reference design for an IoT-enabled energy management system (EMS) in which field sensors, an edge gateway, time-series storage, short-term forecasting and a constrained supervisory controller operate as one system. Local inverter and battery protection remain independent, while the supervisory controller issues dispatch set-points every 15 minutes by solving a mixed-integer linear program (MILP) with power-balance, state-of-charge, charge/discharge and load-priority constraints. A hybrid forecaster, combining gradient-boosted trees with a persistence fallback, is benchmarked against three baselines on a 12-month synthetic-but-realistic dataset. For a 100 kWp PV, 300 kWh battery and 75 kW inverter reference case, the proposed EMS reduces unserved energy by 23%, PV curtailment by 31% and battery throughput by 14% relative to a rule-based baseline, while keeping state-of-charge reserve compliance above 98%. MQTT over TLS carries telemetry, and cybersecurity controls follow NISTIR 7628. Data-driven methods complement, but do not replace, electrical protection and operator judgement.

Isah Bello, Ahmad Muhammad Bello, Bashir Yunus Kolapo, Eeman Sani Abdu, Lukman Salihu, "An IoT-Based Supervisory Energy Management System for Hybrid Solar–Battery Mini-Grids with Lightweight Forecasting" International Journal of Scientific Research and Engineering Development, V9(5): Page(1516-1522) September - October 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.