TY  - JOUR
T1  - Fault Detection and Diagnosis of Steel Refining Process Based on Multi Neural Network
AU - , Y. Selaimia AU - , A. Loudjani AU - , H.A. Abbassi 
JO  - International Journal of Soft Computing
VL  - 1
IS  - 2
SP  - 149
EP  - 154
PY  - 2006
DA  - 2001/08/19
SN  - 1816-9503
DO  - ijscomp.2006.149.154
UR  - https://makhillpublications.co/view-article.php?doi=ijscomp.2006.149.154
KW  - Fault detection and diagnosis
KW  -multi neural network
KW  -Radial Basis Function (RBF) neural network
KW  -refining process
AB  - For steel refining process some of parameters are very critical and can induce to a loss of production
by the need of additional corrections in the shape of reblowing. Among these parameters: carbon and
manganese contents, temperature of the final product. In order to monitor such a system, we propose a multi
neural network based fault detection and diagnosis scheme. A serial/parallel homogeneous configuration is
adopted as the basic structure of the detection system. The first stage allows the classification of the sample
according to the nature of the steel nuance to be produced while the second stage of the network allows the
identification of the volume of oxygen necessary to fusion and will be used as an input for the last stage witch
detect and diagnoses the faults. The simulation results illustrated that after training of the neural networks, the
system is successfully detects and diagnoses the different failures.
ER  -