EKF and ANFIS estimator design in multicomponent batch distillation columns

Yildiz U., Güner E., ÖZGEN C., LEBLEBİCİOĞLU M. K.

7th IFAC Symposium on Dynamics and Control of Process Systems, DYCOPS 2004, Cambridge, Amerika Birleşik Devletleri, 5 - 07 Temmuz 2004, cilt.37, ss.631-636 identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Cilt numarası: 37
  • Basıldığı Şehir: Cambridge
  • Basıldığı Ülke: Amerika Birleşik Devletleri
  • Sayfa Sayıları: ss.631-636
  • Anahtar Kelimeler: Adaptive Neurc-Fuzzy Inference System, Batch Distillation, Kalman Filter, Simulation, State Observer
  • Orta Doğu Teknik Üniversitesi Adresli: Evet

Özet

In the control of batch distillation columns, one of the problems is the difficulty of monitoring the compositions. This problem can be handled by estimating the compositions from readily available online temperature measurements using a state estimator. In this study, an extended Kalman Filter (EKF) and an Adaptive Neuro-Fuzzy Inference System (ANFIS) state estimators that infer the product composition in a multicomponent batch distillation column (MBDC) from the temperature measurements are designed and tested using a batch column simulation. The designed EKF and ANFIS estimators are successfully used in the composition - feedback inferential control of MBDC operated under variable reflux-ratio policy with an acceptable deviation from the desired purity level of the products.