Adaptive Random Search Method + Genetic Algorithms for reaction kinetics modelling: CO2 absorption systems

Carol Toro Molina (1) , Rodrigo Rivera-Tinoco (2) , Chakib Bouallou (1)

1 CEP/Paris
2 INOSS Group S.A.

This paper presents a novel hybrid algorithm based on the Adaptive Random Search Method (ARSM) and enhanced with Genetic Algorithms (GA) theory. This algorithm is first validated by modelling Carbonyl sulfide (COS) absorption by N-methyldiethanolamine (MDEA) aqueous solutions. Results show an excellent fit between modelled values and experimental data. Then, the method is tested to model CO2 absorption by NH3 aqueous solutions. In general, results show that the proposed method is faster and slightly more accurate than classic methods such as Downhill simplex, and simple GA.

Mots clés: Absorption system; Adaptive random search; Aqueous solutions; Carbonyl sulfide; Experimental data; Hybrid algorithms; Kinetics modelling; N-methyldiethanolamine; Simple GA; Absorption; Carbon dioxide; Reaction kinetics; Solutions; Genetic algorithms

Carol Toro Molina, Rodrigo Rivera-Tinoco, Chakib Bouallou. Adaptive Random Search Method + Genetic Algorithms for reaction kinetics modelling: CO2 absorption systems. Chemical Engineering Transactions, 2011, 25, pp.19-24. ⟨10.3303/CET1125004⟩. ⟨hal-00660636⟩