Parameter extraction of PV models under varying meteorological conditions using a modified electric eel foraging optimization algorithm

Abstract

Abstract The dependence on photovoltaic (PV) solar systems has increased dramatically to cover the increasing progress of world energy demand. Therefore, accurately specifying the parameters of PV modules is essential for evaluating the behavior and impact of integrating PV systems into electrical systems. In this context, a modified electric eel foraging optimization (MEEFO) is suggested for determining the parameters of solar PV modules. The proposed technique incorporates three improvement strategies: the fitness distance balance (FDB) strategy, fractional-order calculus (FOC), and quasiopposition-based learning (QOBL). These strategies enhance both exploitation and exploration capabilities while helping to prevent local optimization and premature convergence commonly observed in traditional EEFO. First, the proposed MEEFO is evaluated via two benchmark functions, including the basic and CEC 2019 benchmark functions. The results are then compared with those of other novel methods in terms of accuracy, convergence characteristics, and overall performance. The suggested MMEFO is then employed to identify the parameters for the single, double, and triple diode models of various PV cells/modules, including R.T.C. France, PVM752, STM6-40/36, PWP-201, and STP6-120/36. In addition, various meteorological data, such as changes in radiation and temperature, exist. The simulation findings demonstrate that MEEFO outperforms other techniques and is a reliable and superior method for accurately estimating PV module parameters. The application of MEEFO yields the lowest root mean square error (RMSE) values for the considered single, double, and triple diode models of R.T.C. France. Similarly, for STP6-120/36, the RMSE values are 1.660060E−02, 1.66006E−02, and 1.66089E−02, respectively. Additionally, for PWP-20, the RMSE values are 2.425075E−03, 2.42511E−03, and 2.42510E−03, respectively.