Multi-objective optimization of an experimental integrated thermochemical cycle of hydrogen production with an artificial neural network

Abstract In this study, an experimental lab-scale copper-chlorine (Cu–Cl) cycle of hydrogen production is examined and optimized in terms of exergy efficiency and operational costs of produced hydrogen. The integrated process is modeled and simulated in Aspen Plus incorporating the reaction kinetic parameters with a sensitivity analysis of a range of operating conditions. An artificial neural network (ANN) method with machine learning is used to generate a mathematical function that is optimized based on a multi-objective genetic algorithm (MOGA) method. A sensitivity analysis of variations of each design parameter for both the objective functions and the effectiveness of exergy performance relative to operational costs of produced hydrogen is demonstrated. The sensitivity analysis and optimization results are presented and discussed.