In this section
@inbook{Borghesan:2018:10.1016/B978-0-444-64241-7.50100-2,
author = {Borghesan, F and Chioua, M and Thornhill, NF},
booktitle = {Computer Aided Chemical Engineering},
doi = {10.1016/B978-0-444-64241-7.50100-2},
pages = {631--636},
title = {Forecast of persistent disturbances using k-nearest neighbour methods},
url = {http://dx.doi.org/10.1016/B978-0-444-64241-7.50100-2},
year = {2018}
}
TY - CHAP
AB - This paper focuses on the prediction of persistent disturbances based on their past measurements using two versions of the k-nearest neighbours method: an unweighted and a weighted version. Results of tests on data from a refinery show that the two methods can predict the future trend of a disturbance. They also show that the weighted version is more robust against the choice of the number of nearest neighbours used. The method opens up the possibility of model-free feedforward control without the constraint of causality based on the whole history of a measurement.
AU - Borghesan,F
AU - Chioua,M
AU - Thornhill,NF
DO - 10.1016/B978-0-444-64241-7.50100-2
EP - 636
PY - 2018///
SP - 631
TI - Forecast of persistent disturbances using k-nearest neighbour methods
T1 - Computer Aided Chemical Engineering
UR - http://dx.doi.org/10.1016/B978-0-444-64241-7.50100-2
ER -
Nina Thornhill, ABB/RAEng Professor of Process Automation
Centre for Process Systems Engineering
Department of Chemical Engineering
Imperial College London
South Kensington Campus, London SW7 2AZ
Tel: +44 (0)20 7594 6622
Email: n.thornhill@imperial.ac.uk