Optimized Smart Home Energy Management Systems Based on Hybrid Method

Abstract

This paper introduces a dynamic and innovative approach to energy management aimed at cost reduction and enhancing grid stability. The proposed system integrates Tiny Machine Learning (TinyML) and load forecasting into appliance scheduling and the real-time monitoring and control of building energy consumption. Unlike existing energy management systems that marginalize user involvement, this research proposes an edge-based, semi-automated Smart Home Energy Management System (SHEMS). This system facilitates seamless interaction between energy users and their smart environment, enabling load scheduling and efficient control. The adoption of edge computing in SHEMS minimizes the high latency issues associated with cloud computing, ensuring prompt responses. Motivated by goals such as grid stability, energy sustainability, and remote-control capabilities for all controllable loads, SHEMS addresses critical challenges like energy saving and peak load shaving. A case study demonstrated the system’s economic and functional viability, highlighting its capability to reduce energy and operational costs, respond effectively to user commands, and perform computations efficiently. Comparative analysis revealed that implementing SHEMS in a smart building achieved approximately 24% energy savings compared to traditional manual operations. This study underscores the transformative potential of SHEMS in advancing smart building energy management.