The aim of this study is to predict the performance of proton exchange membrane (PEM) fuel cells composed of various cathod materials by applying Neural network modeling. To predict PEM fuel cell performance, cyclic voltammetry charts and electrochemical active surface area (ECSA) calculations are required. At the first stage, required informations were selected based on the experimental data for various cathod materials. Pure platinum, reduced Graphene Oxide- Carbon Black, Carbon Black and Graphene Nanoplatelets- Carbon Black are the common materials as cathod. After that, avaiable data were classified based on two variables: total surface area and percentage of platinum in the electrode. Then, the artificial neural network (ANN) code is written for each group. By running the code, the cathode cyclic voltammetry diagram as well as ECSA is obtained and compared with exprimental data. Results showed that ANN code could accurately predict the cathode cyclic voltammetry diagram and ECSA values. The higher amount of ECSA shows better performace of fuel cell.

Keywords