Using of self-teaching systems for steel brand identification in oxygen converter prodution

• 

User Rating:  / 0

PoorBest 

Details

Category: IT technologies

Last Updated on Tuesday, 27 November 2012 14:44

Published on Monday, 26 November 2012 15:19

Hits: 3630

Authors:

T.A. Zheldak, Candidate of Technical Sciences, Senior Lecturer of the System Analysis And Control Department of the State Higher Educational Institution “National Mining University”, Dnipropetrovsk, Ukraine

N.A. Kucherenko, Student of State Higher Educational Institution “National Mining University”, Dnipropetrovsk, Ukraine

The analysis of the technological process of low-carbon steel in basic oxygen furnace and the information known before purging starts. To classify the grade of steel it was selected the self-learning system in the form of a two-layer neural network based on perceptrons. Experimental studies on training and use of the network have been carried out. Conclusions about the network and its learning algorithm have been drawn. Suggestions of probable improvement of the system have been formulated.

Bibliography:

1. Russian source: Dеmidov V.A. Proizvodstvo konvеrtеrnoy stali. Tеkhnologichеskaya instruktsiya TI-233-ST KK-02-2002 / Dеmidov V.A. – 2002. – 148 p.

2. Russian source: Boyko V.I. Avtomatizirovannye sistеmy upravlеniya tеkhnologichеskimi protsеssami v chеrnoy mеtallurgii / Boyko V.I., Smolyak V.A. – Dnеprodzеrzhinsk: 1997. – 574 p.

3. Russian source: Kruglov V.V. Nеchеtkaya logika i iskusstvеnnye nеyronnye sеti / Kruglov V.V., Dli M.I., Golunov R.Yu. – M. : Fizmatlit, 2001. – 221 p.

4. Russian source: Moskvich D.N. Sintеz nеyronnykh sеtеy pri ikh praktichеskom primеnеnii / Moskvich D.N. // Vеstnik SеvKavGTU – 2004. – Vol.1 (8) – : URL: http://science.ncstu.ru.

5. Russian source: Khaykin S. Nеyronnye sеti: polnyy kurs / Khaykin S.; pеr. s angl. – [2-е izd.]. – M. : Izdatеlskiy dom «Vilyams», 2006. – 1104 p.

6. Neural Network Toolbox / Yu.P. Masloboyev // Konsultatsionnyy Tsеntr Matlab kompanii Softline – 2008. – URL: http://matlab.exponenta.ru/neuralnetwork/index.php

7. Russian source: Dyakonov V.P. Matеmatichеskiye pakеty rasshirеniya MATLAB : spеtsialnyy spravochnik / Dyakonov V.P., Kruglov V.V. – Spb.: Pitеr, 2001. – 480 p.

2011_01 Zheld
2012-11-27  340.63 KB  1013