Title :
On nonlinear modelling using velocity-based multiple model networks
Author :
McLoone, Seamus C. ; Irwin, George W.
Author_Institution :
Sch. of Electr. & Electron. Eng., Queen´´s Univ., Belfast, UK
Abstract :
A recently proposed blended multiple model system which employs velocity-based local linear models is applied to a Continuous Stirred Tank Reactor (CSTR) plant. Three important practical issues are investigated: appropriate blending of the local models, obtaining the input derivative and suitable choice of weighting functions. Comparison is made with a conventional multiple model system of affine local models. Simulation experiments illustrate and confirm previous analytical results on the dynamical modelling superiority of the velocity-based local model network. Furthermore, since the underlying local models of the velocity-based system are continuous-time and linear, continuity is now provided with established linear theory
Keywords :
chemical technology; modelling; nonlinear systems; Continuous Stirred Tank Reactor; affine local models; input derivative; multiple model system; nonlinear modelling; velocity-based local linear models; weighting functions; Analytical models; Continuous-stirred tank reactor; Control design; Control system synthesis; Electronic mail; Intelligent control; Intelligent systems; Neural networks; Nonlinear systems; Velocity control;
Conference_Titel :
American Control Conference, 2001. Proceedings of the 2001
Conference_Location :
Arlington, VA
Print_ISBN :
0-7803-6495-3
DOI :
10.1109/ACC.2001.945959