Ivan Miletic

On-line applications of multivariate statistics at Dofasco

Abstract There are many complexities involved in being able to accurately analyze and model production processes in a fully integrated steel facility. The drive for improved steel product quality has contributed to significant upgrades in instrumentation and data acquisition infrastructure with the hope of developing more useful information and better process knowledge. While this increased level of instrumentation has made more data available to the analyst, the associated data analysis and modeling problem has become more complicated due to the large and ever-increasing volume of process and product quality data. A class of technologies that Dofasco has used to support improved process modeling, feedback and statistical control in its automation applications is multivariate statistics with a primary focus on Principal Components Analysis (PCA) and Projection to Latent Structures (PLS). These methods have been successfully applied at Dofasco since 1993 to analyze data for a variety of purposes to develop on-line predictive models and process monitoring systems. Two on-line applications are described to illustrate the application of this technology. The first example is an on-line monitoring system that is used to observe the operation of the casting process at the mould area of Dofasco’s number one continuous slab caster. The second example is a control system for determining the optimal amount of reagent needed to accurately remove sulfur from pig iron.

Start Cast Breakouts Preventative Prediction Using Multi-Way PCA Technology

Abstract Breakouts during continuous caster start-up operations are of major concern in the steel-making industry, because they can lead to severe damage to equipment, significant process downtime, and potential safety consequences. As a multivariate statistical (MVS) analysis tool, Multi-way PCA (MPCA) is applied for monitoring the start-up operation of a continuous caster in order to predict potential start cast breakouts so the caster can be automatically stopped to avoid the catastrophic event. It is shown to provide good prediction of start cast breakouts resulting in significant savings in operating and maintenance costs. An on-line start cast monitoring system has been successfully implemented at Dofascos #2 continuous caster.

Applications of multivariate statistics at Dofasco

Multivariate statistical technologies, the principal components analysis and projection to latent structures, are data modeling technologies based on advanced multivariable statistical methods. These methods are capable of: analyzing process data; building predictive models and providing SPC functionality by extracting information from all process and quality data from an operation simultaneously. Multivariate statistical methods are especially powerful techniques for analyzing industrial data sets that have the following characteristics: higher dimensionality; high collinearity; noisy; and with some missing data. The application of these methods have been successfully done at Dofasco since 1993 to analyze data for a variety of purposes, develop online predictive models, and develop online process monitoring systems. An online application is described to illustrate the advantages of this technology.