Implementación del método -d-sda y variables de decisión externa- para el diseño óptimo de columnas de destilación extractiva

The optimal design of an extractive distillation column is approached in this study considering both continuous and discrete design variables. First, it was modeled by the general formulation of a MINLP problem using binary decision variables. Next, three reformulations of the problem were carried o...

Full description

Autores:
Montaña Rosado, Iván Darío
Tipo de recurso:
Trabajo de grado de pregrado
Fecha de publicación:
2021
Institución:
Universidad de los Andes
Repositorio:
Séneca: repositorio Uniandes
Idioma:
spa
OAI Identifier:
oai:repositorio.uniandes.edu.co:1992/51656
Acceso en línea:
http://hdl.handle.net/1992/51656
Palabra clave:
Destilación extractiva
Destilación
Optimización matemática
Columnas de destilación
Variables (Matemáticas)
Ingeniería
Rights
openAccess
License
http://creativecommons.org/licenses/by-nc-nd/4.0/
Description
Summary:The optimal design of an extractive distillation column is approached in this study considering both continuous and discrete design variables. First, it was modeled by the general formulation of a MINLP problem using binary decision variables. Next, three reformulations of the problem were carried out, expressing the binary variables as a function of external variables. Later, the problem was solved using the D-SDA algorithm, which allows the problem to be separated into a master-IPLC that varies the discrete variables, and an FNLP subproblem that optimizes the continuous variables in each discrete step. Finally, the performance of the different possible neighborhoods for the master-IPLC and of the reformulations proposed was evaluated, as well as the efficiency of this method in comparison with other resolution methods, both NLP and MINLP.