Comparison of four methods for parameter estimation of mono- and multi-junction photovoltaic devices using experimental data

The present work analyses four methods used to estimate the physical properties of photovoltaic devices for a single diode model. Two of the most efficient photovoltaic technologies-mono- and multi-junction devices-are used under different temperature and solar radiation conditions for comparing the applicability of each method. Three of the four parameter estimation methods are analytic and the remaining one uses an algorithm for the optimization of non-linear problems, i.e., the generalized reduced gradient. The different methods are summarized and a comparative analysis is performed using experimental data obtained from the literature, highlighting the advantages and disadvantages of each method. Criteria such as the mean absolute percentage error, the coefficient of determination, the absolute error in current calculated at the maximum power point, and computational cost are used. Accordingly, it is concluded that, for all the methods considered in this study, the best accuracy is obtained from simulations using the method proposed by Blas et al. [34] applied to mono-junction modules, and the method proposed by Xiao et al. [39] to multi-junction devices.