modeling and optimization of a hydrogen production unit with artificial neural network and genetic algorithm

The main purpose of this study is to model an industrial unit of hydrogen production based on the conversion of methane to water vapor using an artificial neural network. the prediction of these two factors was considered. highly accurate modeling results predicted absolute mean error, relative mean error. the possible error between actual factory and model data to be 2.14, 1.21, and 2.9 for the first network and 0.37, 0.84, and 0.55 for the second network, respectively. based on the sensitivity analysis, the temperature of the synthesis gas output from the converter had the greatest effect on hydrogen production and waste gas flow rate as the most influential factor on the unit energy consumption. after unit modeling, a genetic algorithm was used to find the optimal operating conditions. in this way, the gross profit obtained from the process was considered an objective function, and the operational factors were optimized to achieve maximum profit using a genetic algorithm. the genetic algorithm results predicted a profit of $ 42.56 per hour, which is 25% higher than the average unit profit in real life.