Hybrid numerical and non-numerical artificial intelligence approaches in steel-making monitoring and control

The rate of steel-making in electrical furnaces is constantly increasing. The relevant problems of steel-making modelling and control are discussed. Suitable approaches of artificial intelligence are presented and a new advanced steel-making monitoring and control system is introduced. In steel-making production in electric arc furnaces the lack of objective information exists. For these reasons it will be effective to use the subjective one and to apply some non-intensive artificial intelligence calculi, namely fuzzy modeling, qualitative modelling, semi-qualitative modeling, neuro-computing and genetic algorithms utilization. Collectively, these methodologies may be viewed as the principal constituents of soft computing with higher tractability, robustness and ease of implementation. The fuzzy-neural and fuzzy-neural-genetic technologies are used for these reasons. To achieve a high level of steel-making process monitoring and control a new advanced steel-making automatic control intelligent system (SACIS) was proposed.