Parallel AC Link Power Converter for Hybrid Renewable Energy Systems Using Machine Learning Control

This paper presents an innovative machine learning-based control method for a power converter system designed to optimize the integration of renewable energy sources, including multi-location photovoltaic (PV) systems and wind turbines (WT), with three-phase AC loads. Unlike conventional power converters, where control strategies often struggle with managing the dynamic variability of renewable energy sources and the fluctuating demands of diverse AC and DC loads, the proposed approach offers a groundbreaking solution. The control method intelligently determines which energy source to prioritize for charging based on its maximum available voltage and which load to prioritize for discharging based on the lowest voltage demand. This selective charging and discharging process is executed in real-time, enabling efficient power distribution regardless of whether the sources are AC or DC. By leveraging machine learning, the system adapts to the inherent variability of renewable sources and load demands, ensuring both fast response times and optimal energy utilization. Simulation results demonstrate the effectiveness of this control strategy, highlighting its potential to enhance control precision and operational efficiency in complex renewable energy systems.