Fuzzy control of semi-batch polymerization reactor with genetic algorithm Yari-kesi̇kli̇ sti̇ren poli̇meri̇zasyon reaktörünün geneti̇k algori̇tma i̇le bulanik kontrolu


ALTINTEN A., ALTINDAL Ş., Erdoǧan S., HAPOĞLU H.

Journal of the Faculty of Engineering and Architecture of Gazi University, cilt.26, sa.3, ss.613-621, 2011 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 26 Sayı: 3
  • Basım Tarihi: 2011
  • Dergi Adı: Journal of the Faculty of Engineering and Architecture of Gazi University
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Sayfa Sayıları: ss.613-621
  • Anahtar Kelimeler: Fuzzy control, Genetic algorithm, Semi-batch polymerization reactors
  • Ankara Üniversitesi Adresli: Evet

Özet

In the polymerization, main aim is the achievement of desired quality product. For this reason, control and design is very important at polymerization processes. In this study, polymerization of styrene in a jacketed semibatch reactor was examined. Manipulated variable was the heat Q which was given from electrical heater. Fuzzy control was chosen as the control method in this study. Membership functions and relation matrix which provide the best fuzzy control were obtained with GA. The fitness function for GA was chosen as the integral of the absolute value of error, IAE. In experimental studies, styrene and toluene mixture was fed continuously to the reactor. As the initiator, benzoyl peroxide was used. Fuzzy control method was used to keep the desired temperature set point. Because reactor temperature is increasing as a result of exothermic reaction. During the experiments, the concentration and molecular weight of polymer were measured analytically by taking samples. It is concluded that GA with fuzzy control is a good way to control the temperature of semi-batch styrene polymerization reactor. It was seen that experimental monomer conversion values are higher than the theoretical values, and viscosity average molecular weights which were obtained experimentally are acceptable when compared with numerical average molecular weights.