Multivariate statistical model-based system for monitoring the operation of a continuous caster and detecting the onset of impending breakouts

摘要

A method of monitoring the operation of a continuous caster in which a multivariate statistical model is developed using off-line measurements of process parameters representing normal operation of the continuous caster. On-line process parameters are measured in real time to determine if the caster operation is normal by calculating test statistics on vectors calculated using the multivariate statistical model with input from the on-line process parameters. The method may advantageously be used to predict breakouts in a continuous caster and allows an operator to take corrective action to avoid such breakouts. Corrective action may also be taken automatically.