Process data chemometrics

Abstract

A data analysis technique successfully applied in spectroscopy to reduce a large quantity of data into meaningful information, called chemometrics, is discussed. Data when properly interpreted by statistical data analysis tools and fundamental and heuristic models yield meaningful information. The use of chemometrics as a multivariate analyzer to provide a composite measurement of the state of a chemical process operation is examined. An application of this analyzer on a Du Pont plant is presented, and two measures to detect and identify important process shifts are introduced.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>