Extraction of double-diode photovoltaic module model's parameters using hybrid optimization algorithm

Abstract EN

This paper presents seven parameters of double diode model of the photovoltaic module under different weather conditions are extracted using differential development with an integrated mutation per iteration (DEIM) algorithm.

The algorithm is produced by integrating of two other algorithms namely, electromagnetism like (EML) and differential evolution (DE) algorithms.

DEIM enhances the mutation step of the original DE by using the attraction-repulsion principle found in the EML algorithm.

Meanwhile, a novel strategy based on adjusting mutation and crossover rate factors for each iteration is adopted in this paper.

The implemented scheme's success is confirmed by comparing its results with those obtained by techniques cited in the literature.

Along with the results, the DEIM suggests high closeness with the experimental I–V characteristic.

For the proposed algorithm an average Root Mean Square Error (RMSE), olute error (AE), mean bias Error (MBE), and execution time values were 0.0608, 0.044, 0.0053 and 23.333, respectively.

The comparisons and evaluations results proved that the DEIM is better in terms of precision and rapid convergence.

Furthermore, fewer control parameters are needed in comparison to EML and DE This paper presents seven parameters of double diode model of the photovoltaic module under different weather conditions are extracted using differential development with an integrated mutation per iteration (DEIM) algorithm.

The algorithm is produced by integrating of two other algorithms namely, electromagnetism like (EML) and differential evolution (DE) algorithms.

DEIM enhances the mutation step of the original DE by using the attraction-repulsion principle found in the EML algorithm.

Meanwhile, a novel strategy based on adjusting mutation and crossover rate factors for each iteration is adopted in this paper.

The implemented scheme's success is confirmed by comparing its results with those obtained by techniques cited in the literature.

Along with the results, the DEIM suggests high closeness with the experimental I–V characteristic.

For the proposed algorithm an average Root Mean Square Error (RMSE), olute error (AE), mean bias Error (MBE), and execution time values were 0.0608, 0.044, 0.0053 and 23.333, respectively.

The comparisons and evaluations results proved that the DEIM is better in terms of precision and rapid convergence.

Furthermore, fewer control parameters are needed in comparison to EML and DE algorithms.