Improving Power Precision in Hybrid Renewable Energy Systems with a Quad Active Bridge DC-DC Converter and Neural Network-Based Decoupling

This paper presents the outcomes of our recent revision on the configurations and controls of hybrid renewable energy systems, incorporating solar panels, a wind turbine, a battery, and a load. We utilize a DC-DC converter (quad active bridge) with a high-frequency transformer to interconnect these sources. This structure reduces the need for multiple voltage adaptation converters and enhances source protection through magnetic coupling. Control is achieved by manipulating phase shifts between transformer voltages to adjust power production based on load demands. Additionally, we introduce a neural network block derived from the real system to dynamically recalculate phase shifts and achieve desired power levels in each source or load, enabling open-loop system control.