Modeling and Intelligent Power Flow Management of a High-Gain Three-Port Converter

The Three-port converter is an electronic power interface that enables simultaneous energy exchange among multiple energy sources and loads. This paper presents the modeling and intelligent control of a High Gain Boost Three-Port Converter (HGBTPC) for dynamic power flow management between the three ports. HGBTPC integrates Photovoltaic (PV) and battery sources to ensure a reliable power supply under varying conditions of source and load. A detailed state-space model of the HGBTPC was developed to capture the converter's dynamic behavior. In addition, a regression neural network was trained with inputs such as PV power, battery State of Charge (SOC), and load demand to predict optimal operating modes in real time. Key validation metrics, such as a confusion matrix, training vs loss accuracy, and mode transition tracking, confirm the effectiveness of the proposed model and the control scheme.