Smart plug-based home energy management with IoT and anomaly detection

This study presents an IoT-based Home Energy Management System (HEMS) using smart plugs and regression-based anomaly detection to optimize household electricity usage. Unlike traditional smart meters, this cost-effective approach provides real-time monitoring and predictive analytics, enabling energy efficiency and safety enhancements. The proposed system integrates Zigbee and Message Queuing Telemetry Transport (MQTT) protocols for seamless communication and employs regression models to identify correlations between power consumption, temperature, and usage duration. Anomaly detection mechanisms identify abnormal consumption patterns, prompting timely interventions. Experimental validation demonstrates high predictive accuracy (R2 > 0.95) and effective anomaly detection. The system offers cost-efficient energy monitoring, improved detection capabilities, and scalable deployment. Future research will explore advanced machine learning techniques and expanded datasets to enhance predictive performance. This study advances smart home energy management by offering a practical and accessible solution for real-time energy monitoring and anomaly detection.