Multivariate Process Monitoring - A method of describing and
signallingprocess behaviour
Ross Sparks and Allan Adolphson, Ross.Sparks@cmis.csiro.au,
Allan.Adolphson@cmis.csiro.au
CSIRO: Mathematical and Information Sciences,
Wednesday 21 October at 2pm (note: special date and time).
Abstract
In this talk, we present methods for monitoring multivariate process
data based on dimension reduction techniques for visualising the data,
while still signalling strange behaviour using all dimensions (all the
data). In contrast to existing methods that are based on static
dimension reduction methodology using a historical training data set, we
dynamically choose the dimensional space using local data.This approach
allows us to detect and describe changes in location, variation, and
correlational structure accurately yet display a large amount of
information concisely. We will illustrate the methodology using examples
of industrial data and also hopefully discuss some of the issues related
to a practical implementation of the method.
Back to HAIL Home Page
Back to Home Page