TY - JOUR
TI - Computationally intelligent modeling and control of fluidized bed combustion process
AU - Ćojbašić Žarko M
AU - Nikolić Vlastimir D
AU - Ćirić Ivan T
AU - Ćojbašić Ljubica R
JN - Thermal Science
PY - 2011
VL - 15
IS - 2
SP - 321
EP - 338
PT - Article
AB - In this paper modelling and control approaches for fluidized bed combustion  process have been considered, that are based on the use of computational  intelligence. Proposed adaptive neuro-fuzzy-genetic modeling and intelligent  control strategies provide for efficient combining of available expert  knowledge with experimental data. Firstly, based on the qualitative  information on the desulphurization process, models of the SO2 emission in  fluidized bed combustion have been developed, which provides for economical  and efficient reduction of SO2 in FBC by estimation of optimal process  parameters and by design of intelligent control systems based on defined  emission models. Also, efficient fuzzy nonlinear FBC process modelling  strategy by combining several linearized combustion models has been  presented. Finally, fuzzy and conventional process control systems for fuel  flow and primary air flow regulation based on developed models and optimized  by genetic algorithms have also been developed. Obtained results indicate  that computationally intelligent approach can be successfully applied for  modelling and control of complex fluidized bed combustion process.