A HYBRID AI-BASED CONTROL FRAMEWORK WITH PSO OPTIMIZATION FOR ENHANCING THE STABILITY AND EFFICIENCY OF RENEWABLE ENERGY IN AUTONOMOUS POWER GRIDS

The integration of renewable energy sources (RES) into autonomous power grids poses significant challenges due to their intermittent, nonlinear, and unpredictable nature. This study presents a hybrid artificial intelligence (AI)-based control system designed to enhance the efficiency, stability, and adaptability of RES in autonomous microgrid environments. The proposed control framework integrates machine learning (ML)-based load forecasting with Particle Swarm Optimization (PSO)-tuned proportional-integral (PI) controllers to dynamically optimize energy flow, voltage regulation, and load balancing. Simulation results, validated using MATLAB®/Simulink®, demonstrate that the hybrid AI-PSO controller significantly outperforms conventional PI control in key performance metrics. Specifically, it reduces voltage deviation from ±6.5 V to ±1.5 V, cuts settling time from 1.2 s to 0.45 s, lowers total harmonic distortion (THD) from 4.5% to 1.8%, and minimizes steady-state error from 2.8 V to 0.6 V. The system also shows enhanced adaptability, rapid compensation of load perturbations, and predictive control of battery energy storage. These results confirm the proposed method’s ability to provide reliable, real-time control under dynamic conditions, supporting its potential application in resilient and sustainable off-grid power systems.