Heats and input variables selection for designing a water detection framework applicable to industrial electric arc furnaces

Abstract

AbstractThis paper describes the development of “heats” and input variables selection models that are incorporated into a water detection framework for an industrial steelmaking electric arc furnace (EAF). The selection models in this work are developed based on latent variable methods. The latent variable methods used in this work are multiway principal component analysis (MPCA) and multiway projection to latent structures (MPLS). The particular problems related to latent variable methods discussed in this paper include data preprocessing, including alignment, unfolding method, centering, and scaling. The outcome of the heats selection model is heats with normal operation and the outcome of the input variables selection model is variables that are highly correlated with the off‐gas water vapour. The water detection framework and developed models are useful in practical settings for the prediction of water leakage and the design of appropriate fault detection and diagnosis strategies.