A real-time appliance monitoring approach with anomaly detection for residential houses

Abstract

Monitoring electrical appliances in residential buildings is essential for minimizing energy waste and enhancing safety through the early detection of abnormal conditions. While researchers have investigated both intrusive and nonintrusive load monitoring approaches, the non-intrusive approach has emerged as preferred due to its cost-effectiveness and noninvasive implementation. Despite considerable progress in appliance monitoring and fault detection systems over the past two decades, critical challenges and limitations persist. This paper proposes a low-complexity appliance identification and monitoring solution to overcome those issues. Furthermore, the proposed solution is integrated with an abnormal condition detection mechanism for critical appliances, aiming to save energy and ensure the safety of the power system. Furthermore, the solution incorporates user feedback via a dedicated mobile application, enhancing adaptability and performance. The proposed solution has been validated in real-time environments using both custom and publicly available datasets, demonstrating improved accuracy in energy monitoring and increased consumer safety.