Premium
Struktursynthese und Optimierung nichtidealer Rektifizierprozesse
Author(s) -
Bauer Markus H.,
Stichlmair Johann
Publication year - 1996
Publication title -
chemie ingenieur technik
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.365
H-Index - 36
eISSN - 1522-2640
pISSN - 0009-286X
DOI - 10.1002/cite.330680803
Subject(s) - convergence (economics) , mathematical optimization , reduction (mathematics) , heuristic , basis (linear algebra) , computer science , scaling , process (computing) , separation (statistics) , mathematics , geometry , machine learning , economics , economic growth , operating system
Optimization processes with MINLP are inherently capable of dramatically reducing the effort and time required for synthesis and analysis of separation processes. This approach represents an extremely powerful tool whose use, however, requires an understanding of process engineering. The principal advantage of the method lies in the simultaneous optimization of both the structure and also the operating conditions of a process. Reliable convergence to the global optimum requires careful limitation, scaling, and prior assignment of all variables. Starting values can be obtained by thermodynamic analysis of the phase equilibrium and by an overall balancing of the individual separation steps. Heuristic reduction of the superstructure generated on the basis of the preferred separation, appears appropriate given the present status of available models and solution algorithms. Under these conditions, the outer approximation algorithm recommended for this model will find the optimum solution after just a few iteration steps.