Premium
Synthesis approach to the determination of optimal waste blends under uncertainty
Author(s) -
Chaudhuri Prosenjit,
Diwekar Urmila
Publication year - 1999
Publication title -
aiche journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.958
H-Index - 167
eISSN - 1547-5905
pISSN - 0001-1541
DOI - 10.1002/aic.690450807
Subject(s) - mathematical optimization , probabilistic logic , parametric statistics , stochastic optimization , computer science , simulated annealing , benchmark (surveying) , sampling (signal processing) , process (computing) , optimization problem , mathematics , artificial intelligence , statistics , geodesy , filter (signal processing) , computer vision , geography , operating system
Abstract The generalized approach to the problem of synthesis under uncertainty is to formulate it as a stochastic optimization problem that involves optimization of a probabilistic function obtained by sampling over uncertain variables. The computational burden of this approach can be extreme and depends on the sample size used for characterizing the parametric uncertainties. A new and efficient approach for stochastic process synthesis is presented. The goals are achieved through an improved understanding of the sampling phenomena based on the concepts derived from fractal geometry. A new algorithm for stochastic optimization based on these concepts to accelerate the process of synthesis under uncertainty is presented. Apart from the benchmark HDA synthesis problem, a real‐world problem of synthesizing optimal waste blends is analyzed to test the applicability of this novel approach in addressing the general problem of synthesis under uncertainty. The solution of this real‐world large‐scale synthesis problem is presented under uncertainty through the application of the new stochastic annealing algorithm, which takes into consideration novel sampling methods used in probabilistic analysis of process models.