[Back]

L. Böhler, M. Kozek et al.:
"Implementation of a fuzzy model predictive controller for biomass combustion";
Talk: e-nova 2017 Zukunft der Gebäude: digital - dezentral - ökologisch, Pinkafeld; 2016-11-24 - 2016-11-25; in: "enova 2017", L. Böhler (ed.); Leykam, Nachhaltige Technologien (2017), ISBN: 978-3-7011-0372-0; 53 - 60.

Implementing state of the art model based control algorithms to supervise and control nonlinear thermo-chemical processes significantly increases their performance. In combination with process feedback from measurements, predictive control algorithms can make use of the underlying models to provide restricted estimations of the likely future process behavior. This information is beneficially used to provide smooth transient behaviour be-tween different set points with respect to actuator saturations. In order to cover the en-tire operating range of the considered nonlinear system, multiple parallel, local linear predictive controller are set up. The data calculated by those individual controller is convoluted and weighted by membership functions according to a set of fuzzy-rules. The result is once again a nonlinear feedback-system, which can compensate local inaccuracies and considers future reference changes. In this work, a straight forward approach for the implementation of a fuzzy model predictive controller for a given nonlinear process model is shown.

Model Predictive Control, Fuzzy, Biomass, Combustion

https://publik.tuwien.ac.at/files/publik_281646.pdf