A bidirectional four-port DC–DC converter for grid connected and isolated loads of hybrid renewable energy system using hybrid approach

Abstract

Most four-port converters typically enable bidirectional power flow through the low-voltage side battery port, which is used to discharge to the high-voltage side DC-link and charge from energy sources. However, system-level power management is restricted by the DC-link’s absence of bidirectional power transmission. This manuscript proposes a hybrid approach utilizing a four-port DC–DC converter that can operate in isolation and in conjunction with the grid for hybrid renewable energy systems. Moreover, the converter architecture enables bi-directional power flow between all four ports, including the high-voltage DC-link, allowing for flexible and efficient power management. The Random Decision Forest and Jellyfish Search technology are combined to form the JS-RDF technique, which goes by that name. The primary goal of the proposed method is to reduce power losses, enhance system performance, and ensure stable voltage profiles. The JS is used for robust optimization, adapting the converter to various conditions, while the RDF employs machine learning for optimal control pulse prediction, enhancing overall efficiency. The JS-RDF approach is implemented on the MATLAB platform and is compared with existing approaches. Also, the JS-RDF method demonstrates great power compared to other existing approaches. From the result, the proposed technique shows outstanding performance with minimal switching losses at 0.19 W and conduction losses at 0.43 W, leading to the lowest total losses of 0.62 W. This emphasizes the superior efficiency of the proposed approach in optimizing power conversion, highlighting its potential to improve the overall performance of converter systems.