Documents disponibles écrits par cet auteur

Embedded control for optimizing flexible dynamic process performance / Jeonghwa Moon in Industrial & engineering chemistry research, Vol. 50 N° 9 (Mai 2011) 

Public ISBD [article]  in Industrial & engineering chemistry research > Vol. 50 N° 9 (Mai 2011) . - pp. 4993-5004 Titre : Embedded control for optimizing flexible dynamic process performance Type de document : texte imprimé Auteurs : Jeonghwa Moon, Auteur ; Seon Kim, Auteur ; Andreas A. Linninger, Auteur Année de publication : 2011 Article en page(s) : pp. 4993-5004 Note générale : Chimie industrielle Langues : Anglais (eng) Mots-clés : Optimization Résumé : High-performance processes require a design that operates close to design boundaries and specifications, while still guaranteeing robust performance without design constraint violations. In order to safely approach tighter boundaries of process performance, much attention has been devoted to integrating design and control in which dynamic controllability, as well as the design decisions, are considered simultaneously. However, rigorous methods solving design and control simultaneously lead to challenging mathematical formulations that easily become intractable numerically and computationally. This paper introduces a new mathematical formulation to reduce this combinatorial complexity of integrating design and control. We will show that a substantial reduction in problem size can be achieved using embedded control decisions within specific designs. These embedded control decisions avoid a combinatorial explosion of control configuration, using a full state space model that does not require a pairing of control variables and loops. The current capabilities of the methodology will be demonstrated using a realistic reactor-column flowsheet. DEWEY : 660 ISSN : 0888-5885 En ligne : http://cat.inist.fr/?aModele=afficheN&cpsidt=24128629 [article] Embedded control for optimizing flexible dynamic process performance [texte imprimé] / Jeonghwa Moon, Auteur ; Seon Kim, Auteur ; Andreas A. Linninger, Auteur . - 2011 . - pp. 4993-5004. Chimie industrielle Langues : Anglais (eng) in Industrial & engineering chemistry research > Vol. 50 N° 9 (Mai 2011) . - pp. 4993-5004 Mots-clés : Optimization Résumé : High-performance processes require a design that operates close to design boundaries and specifications, while still guaranteeing robust performance without design constraint violations. In order to safely approach tighter boundaries of process performance, much attention has been devoted to integrating design and control in which dynamic controllability, as well as the design decisions, are considered simultaneously. However, rigorous methods solving design and control simultaneously lead to challenging mathematical formulations that easily become intractable numerically and computationally. This paper introduces a new mathematical formulation to reduce this combinatorial complexity of integrating design and control. We will show that a substantial reduction in problem size can be achieved using embedded control decisions within specific designs. These embedded control decisions avoid a combinatorial explosion of control configuration, using a full state space model that does not require a pairing of control variables and loops. The current capabilities of the methodology will be demonstrated using a realistic reactor-column flowsheet. DEWEY : 660 ISSN : 0888-5885 En ligne : http://cat.inist.fr/?aModele=afficheN&cpsidt=24128629

Multiscale modeling and solution multiplicity in catalytic pellet reactors / Kedar Kulkarni ; Jeonghwa Moon ; Libin Zhang in Industrial & engineering chemistry research, Vol. 47 n°22 (Novembre 2008) 

Public ISBD [article]  in Industrial & engineering chemistry research > Vol. 47 n°22 (Novembre 2008) . - p. 8572–8581 Titre : Multiscale modeling and solution multiplicity in catalytic pellet reactors Type de document : texte imprimé Auteurs : Kedar Kulkarni, Auteur ; Jeonghwa Moon, Auteur ; Libin Zhang, Auteur Année de publication : 2008 Article en page(s) : p. 8572–8581 Note générale : Industrial chemistry Langues : Anglais (eng) Mots-clés : Pellet  Reactors Résumé : Transport and reaction phenomena in catalytic pellet reactors are often difficult to analyze because of coupling between heat and mass transport occurring at different space and time scales. To calculate the reactor concentrations and temperatures, it is necessary to account for the species reaction and transport occurring in the reactor bulk at the macroscopic level as well as the catalyst pellets at the microscopic level. The resulting approach yields a large system of nonlinear partial differential equations with multiple scales and solutions that are difficult to find numerically. In addition, the catalyst pellets may operate in multiple steady states for identical conditions. Conventional computational methods may entirely miss the multiplicity phenomenon at the catalyst pellet level and, as a result, may not correctly predict overall reactor yields. In this paper, we introduce two numerical techniques to address multiple scales and multiplicity in heterogeneous reaction models. The first method expands existing bisection with “shooting”; the second global method deploys orthogonal collocation over finite elements with niche evolutionary algorithms. We also propose a new multiscale method entitled effectiveness factor maps to expedite and simplify the numerical effort to solve transport and reaction phenomena at different length scales. En ligne : http://pubs.acs.org/doi/abs/10.1021/ie8003978 [article] Multiscale modeling and solution multiplicity in catalytic pellet reactors [texte imprimé] / Kedar Kulkarni, Auteur ; Jeonghwa Moon, Auteur ; Libin Zhang, Auteur . - 2008 . - p. 8572–8581. Industrial chemistry Langues : Anglais (eng) in Industrial & engineering chemistry research > Vol. 47 n°22 (Novembre 2008) . - p. 8572–8581 Mots-clés : Pellet  Reactors Résumé : Transport and reaction phenomena in catalytic pellet reactors are often difficult to analyze because of coupling between heat and mass transport occurring at different space and time scales. To calculate the reactor concentrations and temperatures, it is necessary to account for the species reaction and transport occurring in the reactor bulk at the macroscopic level as well as the catalyst pellets at the microscopic level. The resulting approach yields a large system of nonlinear partial differential equations with multiple scales and solutions that are difficult to find numerically. In addition, the catalyst pellets may operate in multiple steady states for identical conditions. Conventional computational methods may entirely miss the multiplicity phenomenon at the catalyst pellet level and, as a result, may not correctly predict overall reactor yields. In this paper, we introduce two numerical techniques to address multiple scales and multiplicity in heterogeneous reaction models. The first method expands existing bisection with “shooting”; the second global method deploys orthogonal collocation over finite elements with niche evolutionary algorithms. We also propose a new multiscale method entitled effectiveness factor maps to expedite and simplify the numerical effort to solve transport and reaction phenomena at different length scales. En ligne : http://pubs.acs.org/doi/abs/10.1021/ie8003978

parallel hybrid algorithm for process flexibility analysis / Jeonghwa Moon in Industrial & engineering chemistry research, Vol. 47 n°21 (Novembre 2008) 

Public ISBD [article]  in Industrial & engineering chemistry research > Vol. 47 n°21 (Novembre 2008) . - p. 8324–8336 Titre : parallel hybrid algorithm for process flexibility analysis Type de document : texte imprimé Auteurs : Jeonghwa Moon, Auteur ; Kedar Kulkarni, Auteur ; Libin Zhang, Auteur Année de publication : 2008 Article en page(s) : p. 8324–8336 Note générale : Chemical engineering Langues : Anglais (eng) Mots-clés : Flexibility  analysisHybrid  algorithm Résumé : Flexibility analysis is an important task for the optimal design and synthesis of chemical processes with uncertainty. It is a challenging problem because of the discontinuity and nonconvexity of rigorous flexibility programming formulations. In this article, we propose a new parallel hybrid algorithm based on stochastic search in conjunction with a nearest constraint projection technique to numerically solve the flexibility index problem. The proposed method can be applied regardless of the convexity of the design constraints. The stochastic method robustly identifies the global solution without the need for derivative information. The new nearest constraint projection technique is used to handle the constraints of the flexibility index problem in reduced state space. In contrast to existing methods, this technique does not require the addition of artificial variables for active constraints, does not need to have access to explicit analytical forms of the problem formulation or its derivatives, and does not solve for additional artificial variables. Its implementation is well-suited for parallel computing so that computational time can be dramatically reduced. Five applications illustrate the efficacy of the proposed method. En ligne : http://pubs.acs.org/doi/abs/10.1021/ie800644n [article] parallel hybrid algorithm for process flexibility analysis [texte imprimé] / Jeonghwa Moon, Auteur ; Kedar Kulkarni, Auteur ; Libin Zhang, Auteur . - 2008 . - p. 8324–8336. Chemical engineering Langues : Anglais (eng) in Industrial & engineering chemistry research > Vol. 47 n°21 (Novembre 2008) . - p. 8324–8336 Mots-clés : Flexibility  analysisHybrid  algorithm Résumé : Flexibility analysis is an important task for the optimal design and synthesis of chemical processes with uncertainty. It is a challenging problem because of the discontinuity and nonconvexity of rigorous flexibility programming formulations. In this article, we propose a new parallel hybrid algorithm based on stochastic search in conjunction with a nearest constraint projection technique to numerically solve the flexibility index problem. The proposed method can be applied regardless of the convexity of the design constraints. The stochastic method robustly identifies the global solution without the need for derivative information. The new nearest constraint projection technique is used to handle the constraints of the flexibility index problem in reduced state space. In contrast to existing methods, this technique does not require the addition of artificial variables for active constraints, does not need to have access to explicit analytical forms of the problem formulation or its derivatives, and does not solve for additional artificial variables. Its implementation is well-suited for parallel computing so that computational time can be dramatically reduced. Five applications illustrate the efficacy of the proposed method. En ligne : http://pubs.acs.org/doi/abs/10.1021/ie800644n