@inproceedings{Sharifzadeh:2011,
author = {Sharifzadeh, M and Thornhill, NF},
title = {Optimal controlled variable selection using a nonlinear simulation-optimization framework},
year = {2011}
}

Download

TY  - CPAPER
AB  - In feedback control, controlled variables are those process variables which are measuredand fed back to controllers. Then in the presence of disturbances, controllers by themeans of manipulating the inputs aim to maintain the controlled variables at theirsetpoints. The objectives for the selection of controlled variables can be conflicting andcompeting. These objectives include minimization of (1) economic losses, (2) inputmanipulations, (3) output variations and (4) changes in process states. This researchaims to present a systematic framework for optimal selection of controlled variables.Each of the above-mentioned objectives is defined within a multi-objective function. Inaddition, the reasoning behind the selection of nonlinear steady state model is explained.The proposed methodology is benchmarked on an industrial distillation train.Optimization programming is presented and the paper discusses how the size of theoptimization problem can be reduced by means of engineering insights and addressingthe concerns regarding feasibility of the developed control structure. The methodologyis scalable to large industrial problems, while maintaining its rigour. The results confirmthat a very good trade-off is established between different objectives.
AU  - Sharifzadeh,M
AU  - Thornhill,NF
PY  - 2011///
TI  - Optimal controlled variable selection using a nonlinear simulation-optimization framework
ER  - 

Download