PREDICTIVE MODELING OF MIXED HALIDE PEROVSKITE CELL USING HYBRID L27 OA TAGUCHI-BASED GA-MLR-GA APPROACH

Abstract

Perovskite photovoltaic cell is regarded as an alternative configuration for the conventional photovoltaic cells predominantly due to its high efficiency. In this paper, a predictive modeling using a hybrid L₂₇ orthogonal array (OA) Taguchi-based Grey relational analysis (GRA), multiple linear regression (MLR) and genetic algorithm (GA) was proposed to optimize the device parameters for better overall performance. The Perovskite photovoltaic cell model is initially constructed and simulated using solar cell capacitance simulator (SCAPS). The final results reveal that the proposed hybrid L₂₇ OA Taguchi-based GRA-MLR-GA approach has effectively optimized the device parameters in which SnO2:F thickness, SnO2:F donor density, ZnO thickness, ZnO donor density, CH₃NH₃PbI_3-xCl_x thickness, CH₃NH₃PbI_3-xCl_x donor density, Spiro-OMeTAD thickness and Spiro-OMeTAD acceptor density are predictively tuned at 0.198 μm, 8.973 x 10¹⁸ cm^-3, 0.039 μm, 8.827 x 10¹⁷ cm^-3, 0.386 μm, 1.929 x 10¹³ cm^-3, 0.233 μm and 8.984 x 10¹⁸ cm^-3 respectively. After the predictive modeling, both FF and PCE of the perovskite photovoltaic cell have been improved for ~5.93% and ~5.78% respectively.