abstract = "Model-based approaches to fault detection and
isolation suffer from the inconvenience that in
practice it is often difficult to obtain an accurate
mathematical model of a nonlinear system of interest. A
way out of this problem is to use data-driven
approaches to model the process input-output behaviour.
In this work, a relatively new genetic programming
technique is employed to design a nonlinear observer
which models the juice temperature at the outlet of an
evaporator at the Lublin Sugar Factory in Poland. The
resulting observer is then used to generate a residual
for fault detection.",