%P 368-374
%I Elsevier Ltd
%V 103
%A T. Ganesan
%A I. Elamvazuthi
%A K.Z. Ku Shaari
%A P. Vasant
%T Swarm intelligence and gravitational search algorithm for multi-objective optimization of synthesis gas production
%R 10.1016/j.apenergy.2012.09.059
%D 2013
%L scholars4045
%J Applied Energy
%O cited By 95
%K Artificial intelligence; Carbon dioxide; Carbon monoxide; Chemical industry; Learning algorithms; Methane; Methanol; Particle swarm optimization (PSO); Steam reforming; Synthesis gas; Synthesis gas manufacture, Carbon dioxide reforming; Carbon monoxide selectivity; Comparative studies; Empirical model; Gravitational search algorithms; Higher hydrocarbons; Methane conversions; Multi objective; Multi objective optimizations (MOO); Normal boundary intersections; Performance metrics; Problem formulation; Swarm Intelligence; Syn gas; Synthesis gas production, Multiobjective optimization, algorithm; ammonia; carbon dioxide; carbon monoxide; chemical industry; empirical analysis; gas production; hydrogen; industrial technology; methane; methanol; multiobjective programming; optimization; oxidation; performance assessment
%X In the chemical industry, the production of methanol, ammonia, hydrogen and higher hydrocarbons require synthesis gas (or syn gas). The main three syn gas production methods are carbon dioxide reforming (CRM), steam reforming (SRM) and partial-oxidation of methane (POM). In this work, multi-objective (MO) optimization of the combined CRM and POM was carried out. The empirical model and the MO problem formulation for this combined process were obtained from previous works. The central objectives considered in this problem are methane conversion, carbon monoxide selectivity and the hydrogen to carbon monoxide ratio. The MO nature of the problem was tackled using the Normal Boundary Intersection (NBI) method. Two techniques (Gravitational Search Algorithm (GSA) and Particle Swarm Optimization (PSO)) were then applied in conjunction with the NBI method. The performance of the two algorithms and the quality of the solutions were gauged by using two performance metrics. Comparative studies and results analysis were then carried out on the optimization results. © 2012 Elsevier Ltd.