Battery-Aware Supervisory Energy Management for a Hybrid Renewable-Grid DC System Using ESP32 and IoT Monitoring

This work builds a hardware-validated, deterministic supervisory controller for hybrid renewable-grid DC systems, closing a gap the literature has left open: existing implementations handle voltage-and-power bus validation, battery-condition classification, hysteresis, anti-chattering, and fault tolerance one at a time, but none brings all of them together in a single, experimentally-tested embedded controller. The system checks voltage and power jointly to confirm bus availability, which rules out the false-positive failure that voltage-only schemes are prone to, and it reads signed battery current, not just magnitude, to tell charging apart from active discharge support under otherwise identical bus conditions. The asymmetric 11.5 V/12.6 V grid-entry/grid-exit hysteresis, paired with a 7-second anti-chattering hold, was tested on bench hardware through repeated boundary crossings: it suppressed 14 of 16 candidate relay actuations that arose from voltage fluctuation near the switching threshold, showing the mechanism actually works under load rather than just on paper. Fault-aware mode retention, an independent emergency-isolation relay, and IoT telemetry round out the controller, and all of it held up across five controlled scenarios that reproduced the expected mode-transition and battery-classification sequences. Together, this gives a reproducible, low-cost reference architecture for laboratory-scale hybrid DC microgrid control.