ANN-augmented adaptive droop/PI control for residential hybrid microgrids with IoT monitoring

Abstract In alignment with Saudi Arabia’s Vision 2030 goals for energy diversification, sustainability, and increased renewable energy penetration, this paper presents an ANN-assisted adaptive droop/Proportional–Integral (PI) control strategy for residential hybrid microgrids integrating photovoltaic (PV) systems, optional micro-wind turbines, and battery energy storage. The proposed scheme is designed for a grid-tied residential microgrid with islanding capability and is modeled and simulated in MATLAB/Simulink. By dynamically tuning droop coefficients and PI gains in real time, the controller enhances system adaptability under fluctuating renewable generation and rapidly varying residential loads. Simulation results demonstrate that, compared with conventional fixed-gain drop/PI control, the proposed approach improves renewable energy utilization by approximately 15%, increases PV self-consumption by 20%, and reduces peak grid import by about 18%, while maintaining stable voltage and frequency regulation. Efficient energy shifting is achieved through coordinated battery operation, with charging during mid-day solar peaks and discharging during evening demand; battery power reaches up to 5 kW, and State of Charge (SoC) is maintained within 35–90% limits. An Internet of Things (IoT) monitoring layer implemented using ThingSpeak provides near real-time telemetry, with median communication latency below 500 ms and full command success rate. A detailed case study based on the climatic conditions and AC-dominated residential load profiles of Jeddah, Saudi Arabia confirms reduced PV curtailment, improved self-consumption, and decreased reliance on grid imports during peak demand periods. The results highlight the effectiveness of ANN-driven adaptive control combined with secure IoT monitoring in enhancing the performance and operational resilience of residential microgrids in challenging environments.