Belt speed control in a sintering plant using neural networks

Citation
M. Jang et al., Belt speed control in a sintering plant using neural networks, STEEL RES, 69(10-11), 1998, pp. 398-405
Citations number
13
Language
INGLESE
art.tipo
Article
Categorie Soggetti
Metallurgy
Journal title
STEEL RESEARCH
ISSN journal
0177-4832 → ACNP
Volume
69
Issue
10-11
Year of publication
1998
Pages
398 - 405
Database
ISI
SICI code
0177-4832(199810/11)69:10-11<398:BSCIAS>2.0.ZU;2-V
Abstract
Sintering transforms fine-grained ore into lumped ore so that the latter ca n be used in a blast furnace. The fine-grained ore combined with coke and o ther materials is loaded into a sinter box and moved along by the sintering belt while the ignited coke burns. The speed by which the belt moves deter mines how much sintering takes place. Since the process is complicated and lacks an accurate mathematical model, human operators manually control the speed by monitoring various factors in the plant. In this paper, a neural n etwork-based sintering belt speed controller is proposed which copies human operator knowledge. Actual process data were collected from a sintering pl ant for eight months and preprocessed to remove noisy and inconsistent data . A multilayer perceptron was trained using a backpropagation learning algo rithm. in on-line testing at the sintering plant. the proposed model reliab ly controlled the sintering belt speed during normal operation without the help of human operators. Moreover, the quality and productivity was as good as with human operators.