CATEGORY: DEBUTANIZERS
New York Science Journal 2013;6(12)
ANFIS Approach for Identification of Debutanizer
Column
Hamed Sahraie (1), Ali Ghaffari (2), Majid
Amidpour (2)
Hamed.Sahraie@yahoo.com
Ghaffari@kntu.ac.ir
Amidpour@kntu.ac.ir
1. National Iranian Southfield Oil Co
2. K.N. Toosi University of Technology, Tehran, Iran
Abstract:
Describes use of the Adaptive Neuro-Fuzzy
Inference System (ANFIS) to identify and model an actual debutanizer column in
an Iranian refinery. The outputs of the dynamic model, in addition to recent
inputs, depend on previous outputs and inputs. Selected inputs and outputs are
those that are used as manipulated and controlled variables.
The type and number of membership functions were
obtained using a trial and error approach. The optimal configuration was chosen
by root mean square error (RSME) criterion. According to RMSE between real and
simulated outputs, the obtained model is acceptable with the aim of control.
Free Full Text Source: http://www.sciencepub.net/newyork/ny0612/001_21554ny0612_1_9.pdf
No comments:
Post a Comment