Enhanced regulation and optimization techniques for isolated fourth-order l3c resonant converters in solar pv to battery pack conversions

With the growing demand for electric vehicles (EVs) and the push for renewable energy, efficient charging systems for high-voltage EV battery packs remain a challenge, particularly when integrating variable solar power. Traditional charging systems struggle with low conversion efficiency, high operational costs, and the complexity of managing voltage regulation under changing environmental conditions. As EV adoption increases, there is an urgent need for cost-effective, efficient solutions that can optimize charging performance while adapting to fluctuating solar energy availability. This paper proposes a hybrid approach for isolated L3C resonant converters in charging high-voltage battery banks for EV with integrated solar photovoltaic (SPV) systems. The novelty of this manuscript lies in an innovation of Greater Cane Rat Algorithm (GCRA) and Spatial Bayesian Neural Network (SBNN). Therefore, it is known as GCRA-SBNN. The main goal of this proposed method is to minimize the cost and maximize the overall efficiency of the system. The GCRA approach is used to optimize the performance of solar PV source according to the environmental conditions and the SBNN approach is used to predict the voltage regulation for solar PV to high-voltage battery pack applications. By then, the performance of the proposed strategy is implemented in MATLAB platform and compared to various existing techniques like Adaptive Neuro-Fuzzy Interface System (ANFIS), Artificial Neural Network (ANN), and Space Vector Pulse Width Modulation (SVPWM) algorithm. The existing method shows costs of 240$, 275$, and 325$, while the proposed method is cost at 175$. The existing method achieves efficiencies of 85 %, 75 %, and 62 %, whereas the proposed method has an efficiency of 95 %. This demonstrates that the proposed method offers both higher efficiency and lower cost. Compared to the existing methods, the proposed method is a more cost-effective and efficient solution. Overall, the proposed technique stands out in terms of both performance and affordability.