Adaptive Home Energy Management: Human-Centric RL Approach for Diverse Situations

The increased demand for Smart Home control technologies and the rapid growth of AI-based approaches provide an opportunity to develop systems that improve conveniences and reduce potential barriers to adopting renewable technologies, including Electric Vehicles, Photo Voltaic Solar Panels, and Household Batteries (or Energy Storage Systems). Reinforcement Learning is an AI tool for training systems to perform complex tasks and achieve challenging goals. This paper introduces a human-centric RL-based Home Energy Management System that manages a homeowner’s requirements while maximizing comfort and convenience. The homeowner can alter the system’s behaviour depending on existing situations and needs. The system can control the home’s energy resources in the presence of the uncertainty and variability of solar power generation, the varying electricity demands, and the priorities of a homeowner.