Título

DETECTION OF PROCESS- AND MACHINE ANOMALIES IN CONTINUOUS CASTING BY MONITORING THE STRAND WITHDRAWAL FORCE USING MACHINE LEARNING

DETECTION OF PROCESS- AND MACHINE ANOMALIES IN CONTINUOUS CASTING BY MONITORING THE STRAND WITHDRAWAL FORCE USING MACHINE LEARNING

Autoria

Ramstorfer, Franz; Sato, Gabriel Yoshiharu; Demuner, Leonardo Martins

DOI

10.5151/2594-5327-34813

Downloads

Baixar Artigo 10 Downloads

Resumo

 

IN THE PRESENT WORK, THE DEVELOPMENT OF AN ANOMALY DETECTION METHODOLOGY OF THE STRAND GUIDE OF A CONTINUOUS CASTING MACHINE BASED ON THE MONITORING OF THE STRAND EXTRACTION FORCE IS SHOWN. THE ANOMALY DETECTION IS BASED ON THE COMPARISON OF THE ACTUAL TOTAL DRIVE FORCE WITH THE NOMINAL VALUE, CALCULATED BY AN ARTIFICIAL NEURAL NETWORK (ANN), WHICH WAS TRAINED AND OPTIMIZED USING THE MOST RELEVANT CASTING PARAMETERS OF ACTUAL PRODUCTION DATA. THIS PAPER GIVES AN OVERVIEW ABOUT THE DEVELOPMENT PROCESS, THE GENERAL BEHAVIOR OF THE TOTAL EXTRACTION FORCE AND THE PRACTICAL APPLICATION OF THE SYSTEM IN PREVENTIVE CASTER MAINTENANCE.

Palavras-chave

Continuous slab casting, Anomaly detection, strand withdrawal force, Machine learning

Continuous slab casting, Anomaly detection, strand withdrawal force, Machine learning