A novel step-up modified SEPIC converter to enhance photovoltaic system performance using a hybrid ANN-PSO algorithm

Abstract This paper presents a modified SEPIC converter intended for solar applications and high-voltage devices. This design combines a discrete inductor-circuit (3 diodes and 2 inductors) with a switched capacitor-circuit (2 capacitors and 2 diodes), achieving a voltage gain six times the input voltage at a low commitment cycle of 0.5 without using a double inductance. Solar power systems are subject to fluctuations in production due to temperature and radiation changes, which leads to variable and unstable voltage. The SEPIC rectifier converter converts variable voltage into regular DC voltage suitable for powering loads, enhancing the energy extraction efficiency of solar panels. A simulation was performed using Simulink/MATLAB to test the overall performance of the converter based on the ANN-PSO algorithm, where the modified SEPIC was compared with the conventional converter. The results confirmed the effectiveness of the transformer in extracting energy with 100% efficiency from the PV panels. The transformer has proven its effective performance in dealing with photovoltaic systems. The output voltage control was evaluated using a PI controller based entirely on the ANN-PSO optimization algorithm, Where ANN predicts radiation and temperature and trains and teaches it with the pre-existing parameters in the neural to reach the target while the pso algorithm improves the neural results and achieves the optimal performance to reach the best control. and the simulation results confirmed the effectiveness of the proposed approach.