A Self-Learning Edge AI Framework for Disaster-Resilient Renewable Energy Systems

Dr. Ramya Rani N¹, R. Swaranambigai², Bhavatarini M S³, Boomathi M⁴, Abinaya D⁵, Geetha sree G⁶

---

ABSTRACT

The research evaluates the performance of a Self-Learning Energy Node which uses Edge Artificial Intelligence (AI) for disaster-proof renewable energy systems. The system operates independently during grid outages and flood and cyclone situations because centralized control systems and internet access become unavailable. The edge device uses a lightweight supervised machine learning model to differentiate between normal and disaster system states by monitoring solar generation and battery state-of-charge (SOC) and load demand and grid availability. The proposed architecture establishes dynamic load prioritization which maintains continuous power supply to essential loads while disconnecting non-essential loads during emergency situations. The trained model uses TensorFlow Lite for optimization and operates on ESP32/Raspberry Pi to achieve quick inference results. The experimental results confirm that disaster detection achieves over 85% accuracy with a response time under 2 seconds. The framework uses decentralized edge-based technology to improve reliability while decreasing grid dependency and enhancing emergency energy management capabilities.

Index Terms: Edge AI, Renewable Energy Systems, Disaster Detection, Smart Microgrid, Load Prioritization, TensorFlow Lite.

How to cite this article: Ramya Rani N, Swaranambigai R, Bhavatarini MS, Boomathi M, Abinaya D, Geetha sree G. A Self-Learning Edge AI Framework for Disaster-Resilient Renewable Energy Systems. Int J Drug Deliv Technol. 2026;16(19s): 1-7. DOI: 10.25258/ijddt.16.19s.1

Source of support: Nil.

Conflict of interest: None