Abstract
Genetic Programming (GP) is used to develop
inferential estimation algorithms for two industrial
chemical processes. Within this context, dynamic
modelling procedures (as opposed to static or
steady-state modelling) are often required if accurate
inferential models are to be developed. Thus, a simple
procedure is suggested so that the GP technique may be
used for the development of dynamic process models.
Using measurements from a vacuum distillation column
and an industrial plasticating extrusion process, it is
then demonstrated how the GP methodology can be used to
develop reliable cost effective process models. A
statistical analysis procedure is used to aid in the
assessment of GP algorithm settings and to guide in the
selection of the final model structure.
Users
Please
log in to take part in the discussion (add own reviews or comments).