A novel two - level optimization framework based on constrained ordinal optimization and evolutionary algorithms for scheduling of multipipeline crude oil blending / Bai Liang in Industrial & engineering chemistry research, Vol. 51 N° 26 (Juillet 2012)  

Public ISBD [article]  inIndustrial & engineering chemistry research > Vol. 51 N° 26 (Juillet 2012) . - pp. 9078–9093 Titre : A novel two - level optimization framework based on constrained ordinal optimization and evolutionary algorithms for scheduling of multipipeline crude oil blending Type de document : texte imprimé Auteurs : Bai Liang, Auteur ; Jiang Yongheng, Auteur ; Huang Dexian, Auteur Année de publication : 2012 Article en page(s) : pp. 9078–9093 Note générale : Industrial chemistry Langues : Anglais (eng) Mots-clés : Optimization Oil Résumé : This paper introduces a practical scheduling of multipipeline crude oil blending (SMCOB) problem. It is formulated as a complex mixed integer nonlinear programming (MINLP) model, taking the charging sequence and flow rates of oil tanks as decision variables, which cannot be efficiently solved by traditional deterministic methods and solvers. Then, a novel two-level optimization framework based on constrained ordinal optimization (COO) and evolutionary algorithms (EA) is proposed. The solution methodology has two stages based on the main procedures of COO. At the crude evaluation stage, discrete EA are used to search for sequence solutions in the outer level. It evolves the sequence solutions on the basis of their rough evaluation of the feasibility and objective value obtained from the inner level and keeps certain number of probably best sequence solutions. At the accurate evaluation stage, the probably best sequence solutions kept by the crude evaluation stage are accurately evaluated by inner-level continuous EA. The COO approach ensures that some true good enough sequence and flow rate solutions can be obtained from the accurate evaluation stage with high probability. COO-based EA are compared with mixed-coding EA to verify the framework’s efficiency and robustness. ISSN : 0888-5885 En ligne : http://pubs.acs.org/doi/abs/10.1021/ie202224w [article] A novel two - level optimization framework based on constrained ordinal optimization and evolutionary algorithms for scheduling of multipipeline crude oil blending [texte imprimé] / Bai Liang, Auteur ; Jiang Yongheng, Auteur ; Huang Dexian, Auteur . - 2012 . - pp. 9078–9093. Industrial chemistry Langues : Anglais (eng) in Industrial & engineering chemistry research > Vol. 51 N° 26 (Juillet 2012) . - pp. 9078–9093 Mots-clés : Optimization Oil Résumé : This paper introduces a practical scheduling of multipipeline crude oil blending (SMCOB) problem. It is formulated as a complex mixed integer nonlinear programming (MINLP) model, taking the charging sequence and flow rates of oil tanks as decision variables, which cannot be efficiently solved by traditional deterministic methods and solvers. Then, a novel two-level optimization framework based on constrained ordinal optimization (COO) and evolutionary algorithms (EA) is proposed. The solution methodology has two stages based on the main procedures of COO. At the crude evaluation stage, discrete EA are used to search for sequence solutions in the outer level. It evolves the sequence solutions on the basis of their rough evaluation of the feasibility and objective value obtained from the inner level and keeps certain number of probably best sequence solutions. At the accurate evaluation stage, the probably best sequence solutions kept by the crude evaluation stage are accurately evaluated by inner-level continuous EA. The COO approach ensures that some true good enough sequence and flow rate solutions can be obtained from the accurate evaluation stage with high probability. COO-based EA are compared with mixed-coding EA to verify the framework’s efficiency and robustness. ISSN : 0888-5885 En ligne : http://pubs.acs.org/doi/abs/10.1021/ie202224w

Exemplaires

Code-barres	Cote	Support	Localisation	Section	Disponibilité
aucun exemplaire