The Community for Technology Leaders
Green Image
<p>In iron- and steelmaking, as in other industrial activities, increasingly stringent customer requirements for product quality and regularity lead to tighter control of the manufacturing processes. These requirements must be linked to minimize costs: Dead heads (lost steel at the beginning and end of a band) at each step in manufacturing steel coils must be strictly minimized. Moreover, as the leading European iron- and steelmaker, Sollac (in France) needs to be increasingly innovative and must develop new techniques, especially in the field of artificial intelligence.</p> <p>Indeed, AI has greatly helped the steel industry to face its evolutionary challenges, typically with expert systems and fuzzy logic. Nowadays, artificial neural networks (ANNs) play an increasingly important role in this field. Hence, Sollac decided, in collaboration with CRIN-INRIA (Centre de Recherche en Informatique de Nancy-Institut National de Recherche en Informatique et Automatique), to determine which ANN-based preset method should be added to a temper mill tool. Sollac's 20 years of modeling experience with this machine put it in an ideal position to deeply evaluate the contribution of ANN techniques to this process modeling.</p> <p>In this article, we present an original application, the on-line integration of the presetting of a temper mill (a kind of roll mill) named Skinpass in the Sainte-Agathe plant (Florange, France) with an ANN. We briefly present the ANN techniques and describe the industrial context before detailing the ANN design and its on-line implementation. An important validation phase followed this implementation and led to an interesting adaptation of this work involving hybrid model mixing of a symbolic approach and a connectionist approach.</p>

N. Pican, F. Alexandre and P. Bresson, "Artificial Neural Networks for the Presetting of a Steel Temper Mill," in IEEE Intelligent Systems, vol. 11, no. , pp. 22-27, 1996.
95 ms
(Ver 3.3 (11022016))