Svein Håvard Stokka
Department of Electrical Engineering and Computer Science, University of Stavanger, Norway
Kim André Jakobsen
Department of Electrical Engineering and Computer Science, University of Stavanger, Norway
Kristian Thosen
Department of Electrical Engineering and Computer Science, University of Stavanger, Norway
Tormod Drengstig
Department of Electrical Engineering and Computer Science, University of Stavanger, Norway
Download articlehttp://dx.doi.org/10.3384/ecp17138284Published in: Proceedings of the 58th Conference on Simulation and Modelling (SIMS 58) Reykjavik, Iceland, September 25th – 27th, 2017
Linköping Electronic Conference Proceedings 138:38, p. 284-291
Published: 2017-09-27
ISBN: 978-91-7685-417-4
ISSN: 1650-3686 (print), 1650-3740 (online)
In this paper we make a physiological based dynamic
model of a laboratory scale industrial tank process configured
with an inlet pump and an outlet valve. Thus, the
pump and valve characteristics, and also the functional expression
for the hydrostatic pressure, are fitted to enzymatic
mechanisms and saturable signaling kinetics. The
model is verified against process data and shows good
compliance. Based on the model, we tune and apply a
physiological inspired control structure to control the water
level in the presence of disturbances in the inlet pump
speed. This control structure is further compared to an Icontroller,
and we show that one of the main properties of
physiological control is that the steady state level of the
controlled variable may deviate from the setpoint value.
We continue by showing how to lump this deviation and
the setpoint into a variable setpoint controller. The main
benefit of such a controller is reduced variation in the control
signal (calculated as total variation, TV). Based on
these results, we finally suggest an industrial control structure
where the control signal participates in the continuous
calculation of a variable setpoint.