APPLICATION OF ARTIFICIAL NEURAL NETWORKS IN MODELING OF MANUFACTURED FRONT METALLIZATION CONTACT RESISTANCE FOR SILICON SOLAR CELLS

摘要

This paper presents the application of artificial neural networks for prediction contact resistance of front metallization for silicon solar cells. The influence of the obtained front electrode features on electrical properties of solar cells was estimated. The front electrode of photovoltaic cells was deposited using screen printing (SP) method and next to manufactured by two methods: convectional (1. co-fired in an infrared belt furnace) and unconventional (2. Selective Laser Sintering). Resistance of front electrodes solar cells was investigated using Transmission Line Model (TLM). Artificial neural networks were obtained with the use of Statistica Neural Network by Statsoft. Created artificial neural networks makes possible the easy modelling of contact resistance of manufactured front metallization and allows the better selection of production parameters. The following technological recommendations for the screen printing connected with co-firing and selective laser sintering technology such as optimal paste composition, morphology of the silicon substrate, co-firing temperature and the power and scanning speed of the laser beam to manufacture the front electrode of silicon solar cells were experimentally selected in order to obtain uniformly melted structure well adhered to substrate, of a small front electrode substrate joint resistance value. The prediction possibility of contact resistance of manufactured front metallization is valuable for manufacturers and constructors. It allows preserving the customers' quality requirements and bringing also measurable financial advantages.