EXPERT SYSTEM DESIGN AND CONTROL OF CRUDE OIL DISTILLATION COLUMN OF A NIGERIAN REFINERY USING ARTIFICIAL NEURAL NETWORK MODEL

Popoola, L.T. and Gutti, Babagana and Alfred, Akpoveta Susu (2013) EXPERT SYSTEM DESIGN AND CONTROL OF CRUDE OIL DISTILLATION COLUMN OF A NIGERIAN REFINERY USING ARTIFICIAL NEURAL NETWORK MODEL. International Journal of Research and Reviews in Applied Sciences, 15 (3). pp. 337-346.

[img]
Preview
Text
ijrras_15_3_16.pdf

Download (512kB) | Preview
Official URL: https://arpapress.com

Abstract

This research work investigated the expert system design and control of crude oil distillation column (CODC) using artificial neural network model which was validated using experimental data obtained from functioning crude oil distillation column of Port-Harcourt Refinery, Nigeria. MATLAB program was written for the artificial neural network back-propagation algorithm using the implementation steps of the artificial neural network. Out of the onehundred and thirty (130) experimental data sets obtained, ninety percent (90%) were used for training the network while the remaining ten percent (10%) were used for testing the network to determine its prediction accuracy. The neural network architecture for the design of the crude oil distillation column was fourteen inputs with one hidden layer and seven outputs (14-1-7); and thirteen (13) inputs with one hidden layer and six (6) outputs (13-1-6) for the neural network controller. The accuracies obtained for the design were 94%, 99%, 92%, 93%, 81%, 95% and 90% for temperature at which 100% (T100) of Kerosene, 90% (T90) of Diesel and 10% (T10) of AGO were distilled; and naphtha, kerosene, diesel and AGO flow rates respectively. The maximum relative error between the experimental data and the calculated data obtained from the output variables of the neural network for CODC design was 1.98% error. The accuracies obtained for the neural network controller (NNC) were 98%, 99%, 99%, 93%, 97% and 97% for the stripping steam to main column, LDO stripper, HDO stripper, reflux flow 1, reflux flow 2 and reflux flow 3 respectively. The little deviation between the output variables of the experimental and calculated data for the cases of NNC predictions for reflux flows 1, 2 and 3 resulted from their excessive usage by the PID controller of the refinery considered to meet the product specifications. Hence, artificial neural network model is an effective tool for the design and control of crude oil distillation column.

Item Type: Article
Uncontrolled Keywords: Crude Oil Distillation Column, Control, Artificial Neural Network Model, Architecture, Input and Output Variables, Design, Back-Propagation Algorithm, PID Controller
Subjects: Q Science > Q Science (General)
Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Q Science > QA Mathematics > QA76 Computer software
Q Science > QD Chemistry
T Technology > T Technology (General)
T Technology > TP Chemical technology
Divisions: Faculty of Engineering, Science and Mathematics > School of Electronics and Computer Science
Depositing User: Mr Tayo Okunlola
Date Deposited: 20 May 2016 11:14
Last Modified: 20 May 2016 11:14
URI: http://eprints.abuad.edu.ng/id/eprint/729

Actions (login required)

View Item View Item