Photovoltaic System Parameter Estimation Using Marine Predators Optimization Algorithm Based on Multilayer Perceptron

Abstract

Solar and wind based renewable energy sources are highly variable according to unstable energy sources due to variable meteorological conditions. Providing a stable DC bus voltage to the inverter’s input by eliminating voltage fluctuations in the generated electrical energy is the main concern of researchers and is highly sensitive in the integration of distributed energy sources. In this study, in order to provide stable and sustainable output voltage at the output of a DC/DC converter, parametric simulation studies are carried out according to different input voltage, duty ratio, and switching frequency values of a DC/DC power converter circuit in a system fed with a solar energy source. Using this dataset obtained from a solar energy source, a parameter estimation in the smart grid structure is studied using artificial hybrid prediction intelligence techniques. Grey wolf and marine predators optimization algorithms are integrated to the multilayer perceptron in the stage of developing the hybrid prediction models. Then, the accuracy of the prediction processes is tested and verified with the power electronics circuit software. Thus, a data analysis approach based on parametric simulation studies in smart grid structures is reported in order to give clear ideas to researchers before designing a prototype.