Ali Shaeri
NIOEC, Iran
Posters & Accepted Abstracts: Int J Appl Sci Res Rev
A feed forward and multilayer perceptron artificial neural network (ANN) with back propagation algorithm were applied for prediction of maximum rate of gasoil production in a live hydrocracker unit (HCR). To this aim, two years of operating conditions were gathered from distributed control system (DCS). The simulation of unit by petro-SIM software was done as the first step to find important operational conditions. A three-layer ANN was adopted to predict maximum gasoil production in terms of aforementioned inputs. To find the best fitted ANN structure, 3342 different structures were examined. It was found that the best structure possessed 4, 7 and 1 neurons for the first, second and third layer, respectively. Furthermore, logsig, logsig and purelin were found the best transfer function for first, second and third layer. The best model was extracted and the obtained data were applied to the live HCR unit with capacity of 40000 bbl/day. The yielded optimum data related to ANN such as reactor temperature, reactor pressure and H2/HC mole ration were selected. Applying the ANN data resulted in an increase in the amount of gasoil and save near to $.
E-mail:
shaeriprocess@gmail.com a.shaeri@iau-tnb.ac.ir
Spanish 
Chinese 
Russian 
German 
French 
Japanese 
Portuguese 
Hindi 