Innovative fuzzy reinforcement learning based energy management for smart homes through optimization of renewable energy resources with starfish optimization algorithm

Population growth and economic development have increased the world’s energy consumption, making it more difficult to manage peak loads and lower the cost of home energy management systems (HEMS). This has led to a need for smart, flexible solutions that incorporate renewable resources to improve sustainability and economic efficiency. To optimize power flow in a hybrid renewable energy system (HRES), this study suggests a fuzzy logic-based energy management system (Fuzzy-EMS) that is improved with reinforcement learning and optimized using the Starfish Optimization Algorithm (SFOA). It incorporates solar photovoltaic (PV), wind turbines (WT), battery storage systems (BSS), and electric vehicles (EVs). Adaptive handling of uncertainties in renewable generation and load demand using a Takagi–Sugeno fuzzy reinforcement learning model with triangular membership functions and 81 rules, real-time energy trading with the upstream grid, and a multi-objective framework that balances cost minimization and renewable utilization maximization are among the main contributions. Cost reductions of 35.2%, 23.8%, and 26.43% under fixed pricing, real-time pricing (RTP), and day-ahead pricing (DAP) models, respectively, are examples of how MATLAB simulations outperform well-known techniques. Furthermore, the system outperforms diesel-based systems by lowering operating costs and carbon emissions by 11.87–18.7% and increasing the use of renewable energy by up to 70% in hybrid scenarios, resulting in a net present cost (NPC) of $269,246 and a levelized cost of electricity (LCOE) of $0.281 over a 20-year period.