Slide Background
Slide Background

NEW TOOLS FOR PROCESS CONTROL IN ALUMINIUM INDUSTRIES - PLSR ON XRD RAW DATA

König, U. Norberg, N.

X-ray diffraction (XRD) is a standard tool for process control in aluminium industries. Varying raw material qualities, the use of different fluxes and increasing prices require a better control of processes and a more efficient use of energy. Traditionally quality control of bauxites, alumina and electrolytic baths has relied on calibration based single peak methods or more advanced full pattern techniques. A common method is the Rietveld quantification which uses structural information to predict information from the full XRD pattern using physical models and fitting techniques. Sometimes this approach is stretched to its limits, especially when no realistic physical model is available, or when the model is either too complex or doesn’t fit to reality. In such cases Partial Least-Squares Regression (PLSR) is a powerful method that predicts information directly from raw XRD data without complex modelling of peak shapes. This paper will describe the advantages of using PLSR to monitor process relevant parameters during the production of bauxite, alumina and aluminium.