S. Rashidi - Department of Chemical Engineering, Ferdowsi University of Mashhad, Mashhad, Iran
M.A. Fanaei - Department of Chemical Engineering, Ferdowsi University of Mashhad, Mashhad, Iran
A. Ahmadpour

Carbon nanostructures are famous structures which are used in several industries such as separation, treatment, energy storage (i.e. methane and hydrogen storage), etc. A successful modeling of activated carbon preparation is very important in saving time and money. There are some attempts to achieve the appropriate theoretical modeling of activated carbon preparation but most of them were almost unsuccessful due to the complexity between the input and output variables. In this paper the empirical modeling of activated carbon preparation from Spanish anthracite based on adaptive neuro-fuzzy inference system (ANFIS) is investigated. ANFIS model is established to delineate the relationship between the BET surface area of the prepared activated carbon with initial and operational conditions; agent type, agent ratio, activation temperature, activation time and nitrogen flow. The results show that the selected model have a good accuracy with a coefficient of determination values (R۲) of ۰.۹۸۸۵ and average relative error (ARE) of ۰.۰۰۲۶۸.

Activated Carbon, ANFIS, Carbon nanostructure, Neural Network