Neuro Fuzzy Technique for Steel Process Modeling

Abstract

Steels are composites of carbon and iron. The steel properties extremely rely on the amount of alloying constituents, so their quantities are carefully measured and controlled during its production. The addition of alloying components in the manufacturing of steel is usually made taking into account the expert experience of specialists and in this manner an administrator decides the amount of alloying components to be added to create the steel of a specific kind. Since the ratio of alloying elements in steel manufacturing normally has indefinite nature, Subtractive Clustering method and Adaptive Network based Fuzzy Inference System (ANFIS) is projected in this work for evaluating the amount of alloying elements with reduced computation errors. The optimal fuzzy rules are evaluated using subtractive clustering and ANFIS method which includes the advantages of fuzzy systems and the neural networks is used to adjust the acquired fuzzy rules produced by means of subtractive clustering. The simulation outcomes signify that the proposed methodology may be implemented effectually for making steel.