Premium
Optimal multiloop feedback design using simulated annealing and neural network
Author(s) -
Lin Julian Jeng Lin,
Wong David Shan Hill,
Yu Shuh Woei
Publication year - 1995
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.690410224
Subject(s) - dept , library science , engineering , computer science , chemistry , stereochemistry
[[abstract]]©1995 Wiley - In this study, artificial neural network is utilized directly for feedback control system design. A neural network is trained on-line to learn the complex and highly nonlinear relationship between tuning parameters, process upsets, and the summation of time-weighted absolute error integrals of the quality control loops of a simulated high-purity distillation column. With the learning process directed by simulated annealing, the number of test runs can be greatly reduced. Optimal tuning parameters are obtained using the ANN model. Satisfactory control of both end products of the binary column is maintained with the proposed design technique.[[department]]化學工程學