Produção Nacional
Revisado por pares
COMPARISON BETWEEN LINEAR-ADAPTIVE, LINEAR LOCAL MODEL NETWORK AND NONLINEAR MPC CONTROLLERS
2001; National University of the South; Volume: 31; Issue: 5 Linguagem: Inglês
ISSN
1851-8796
AutoresArgimiro R. Secchi, Luı́s Gustavo Soares Longhi, Jorge Otávio Trierweiler,
Tópico(s)Advanced Data Processing Techniques
ResumoThere exist some situations where a fixed linear controller can no t be used to control a pro cess. In this wor k, three possibilities to deal with these situations are tested and co mpared: (i) nonlinear control, (ii) adaptive linear control and (iii) local linear model network control. The used controllers belong to the MPC (Model Predictive Co ntrol) categ ory due to the great po pularity of this paradigm at the chemical industries. An iso thermal CSTR (Co ntinuo us Stirred Tank Reacto r) processing the classical Van de Vusse reaction is used as the plant. The compariso n between the different controllers is based on the following criteria: closed-loop perfo rmance, easiness to implement and controller reliability. Finally, an experimental application is presented.
Referência(s)