Determining the dependence of solar cell power conversion efficiency on fabrication parameters in inorganic perovskite types by programming gene expression

Abstract

This article is an attempt to review inorganic perovskite solar cells and propose a new predictive model to estimate the power conversion efficiency (PCE) using gene expression programming (GEP). All references used the spin coating method to construct the solar cells. For the perovskite layer, parameters of thickness, bandgap, and mass percentage of repetitive elements of perovskite layers were considered as input variables, and the PCE was selected as the output variable. In the first stage, 29 experiments were extracted from reliable articles. As a first step to ensure the robustness of the selected data, the box plot was used to determine the distribution of operational parameters, and the outlier's data (three data sets) were excluded. Remained data set used to construct models by training (19 data sets) and testing (7 data sets) in GEP. Statistical indices including the absolute fraction of variance (R 2 ), root mean square error (RMSE), root relative squared error (RRSE), mean squared error (MSE), and box plot of variables were used to evaluate the accuracy of the proposed models. Finally, the best model was selected by R 2 = 0.9111, RMSE = 0.0878, RRSE = 0.2995, and MSE = 0.0077. The results showed that GEP can be used as a unique tool for modeling PCE in perovskite solar cells with operational parameters. Sorting out the effect of each parameter on PCA was the next strength of this study, which was performed using sensitivity analysis and showed that bandgap, the mass percentage of cesium (Cs), lead (Pb) and iodine (I) were the most influencing parameters on the PCE from the inorganic perovskite solar cells, respectively. • The perovskite layer on PCE plays an important role in understanding the behavior of these solar cells. • Review inorganic perovskite solar cells and propose a new predictive model to estimate the PCE using GEP. • GEP can be used as a unique tool for modeling PCE in perovskite solar cells with operational parameters. • Sorting out the effect of each parameter on PCE was strength of this study, which was performed using sensitivity analysis. • Bandgap, mass percentage of Cs, Pb and I were the most influencing parameters on PCE from the inorganic perovskite solar cells.