Artificial Neural Network Aided Design of a Stable Co-MgO Catalyst of High-Pressure Dry Reforming of Methane

Dry reforming of methane attracts much attention in order to convert the two greenhouse gases simultaneously to syngas. Preparation parameters of the citric acid method were surveyed to prepare a Co-MgO catalyst with a long life using the design of experiment (DOE), an artificial neural network (ANN), and a grid search (GS). The preparation parameters such as Co loading, amount of citric acid, calcination temperature, and pelletization pressure were determined according to an L9 orthogonal array. After the catalytic activity was measured in a conventional fixed-bed reactor under pressure, a good fitting of the simple power-law equation (SPLE) to the activity change was obtained. The preparation parameters and the resultant SPLE parameters were used for the training of the ANN. The optimum was determined by a GS and verified experimentally to be stable. The combination of SPLE parameters by DOE, ANN, and GS was found to be a useful tool for the development of the catalyst.