Abstract

This study introduces a sophisticated computational model developed in MATLAB to simulate the performance of interdigitated back contact (IBC) solar cells under various environmental conditions. The main objective of this work is to develop and validate a neural network that can predict Iph and Io of an IBC solar cell under varying environmental conditions (temperature and irradiance). The model is based on a deep feedforward neural network comprising two hidden layers with 20 and 10 neurons, respectively. This architecture enables accurate prediction of key electrical parameters, such as photocurrent (Iph) and reverse saturation current (Io), influenced by factors like temperature and irradiance. Trained on synthetic data representing realistic fluctuations in these variables, the neural network significantly outperforms conventional experimental models in predictive accuracy. The simulation analyzes current–voltage (I–V) and power–voltage (P–V) characteristics over a voltage range of 0 to 0.85 V, incorporating temperature-dependent ideality factors to faithfully represent physical behavior. Iterative optimization strategies were implemented to address convergence issues in open-circuit voltage calculations, ensuring realistic voltage levels (approximately 0.6–0.7 V at 100 W/m²). The model demonstrates impressive performance, accurately predicting short-circuit current (Isc), open-circuit voltage (Voc), maximum power point (Pmax), fill factor (FF), and overall efficiency, all validated against expected IBC cell behavior. This predictive framework provides a versatile tool for optimizing the design and performance of IBC solar cells under dynamic operating conditions.