"Chemometric methods for batch process control using NIR
spectroscopy" - Abstract
The thesis investigates new chemometric methods
that could be applied to near-infrared (NIR) spectroscopic data
to monitor and control efficiently industrial batch processes. Monitoring
and controlling processes is a key activity since it guarantees
that the manufactured products will meet the required specifications.
To perform this controI several kinds of analyses exist, among which,
the in-line analysis that makes the analysis directly in the process
line without taking any sample. This thesis focuses exclusively
on methods that can be applied in-line. The Orthogonal Projection
Approach (OPA) is especially studied because of its simplicity and
because of the interpretability of the results. Batch data present
also an n-way structure (typically samples x variables x batches)
and methods able to deal with them are also described in this thesis.
To control processes in-line, models can be used. However, while
modelling, problems can occur, especially about complexity. The
complexity problem is a general problem in chemometrics and means
to deal with it are described. Another important topic of this thesis
is the selection of appropriate data. The selection of relevant
NIR wavelengths is important from an industrial point of view to
decrease the spectra acquisition time, and consequently also the
time needed to obtain predictions via models. Genetic algorithms
are here used for this purpose. Through the whole thesis, an overview
of the use nf OPA to monitor batch processes in an industrial context
is given. Some multivariate statistics based on Principal Components
Analysis (PCA) are also investigated in the OPA context. Complementary
to OPA, a method called STATIS is described to monitor the evolution
in time of a batch. Finally, some limitations of OPA for batch process
data are studied and some solutions are given.
|