| Abstract: |
This paper presents a comprehensive empirical investigation into the design, implementation, and adaptive control of grid-connected smart inverters intended for seamless integration of renewable energy sources (RES) into modern power distribution networks. With the exponential growth of distributed generation from photovoltaic (PV) systems and wind energy conversion systems (WECS), the demand for high-performance, intelligent power electronics interfaces have become critically important. The proposed smart inverter architecture incorporates a hybrid multilevel inverter (MLI) topology coupled with an Artificial Intelligence-enhanced Model Predictive Control (AI-MPC) strategy that achieves total harmonic distortion (THD) of 0.52%, considerably below the IEEE 1547-2018 standard limit of 5%. Experimental data collected over a 12-month monitoring period at a 25 kW grid-connected pilot installation in Chhattisgarh demonstrates that the proposed system achieves an average MPPT efficiency of 99.2% for solar PV and 97.8% for wind integration, with an overall power conversion efficiency of 97.4%. The system successfully demonstrated Low Voltage Ride-Through (LVRT) compliance across all simulated grid fault scenarios. |
