Home > Archive > 2015 > Volume 5 Number 5 (Oct. 2015) >
IJMLC 2015 Vol.5(5): 388-391 ISSN: 2010-3700
DOI: 10.7763/IJMLC.2015.V5.539

A Neural Network Based Soft Sensor for Online Vapor Product Quality Estimation of a Refinery Debutanizer Column

Bordin Wanichodom, Nont Neamsuwan, and Pornchai Bumroongsri

Abstract—In order to keep crude oil refining products within the specifications, online monitoring and laboratory testing are usually required. Time delay in process monitoring and control may occur since the products from distillation columns must be analyzed in the laboratory. To overcome this problem, a neural network based soft sensor for online measurement of product quality was developed in this paper. A refinery debutanizer was chosen as a study case. Various structures of neural networks with different numbers of neurons in each hidden layer were created and tested for their performance on the estimation of propane composition in the distillate stream. The simulation results showed that the neural network containing 5 and 10 neurons in the first and second hidden layers, respectively, gave the best performance as compared to real industrial data.

Index Terms—Artificial neural network, debutanizer column, online soft sensor, vapor product quality estimation.

Bordin Wanichodom, Nont Neamsuwan, and Pornchai Bumroongsri are with the Department of Chemical Engineering, Faculty of Engineering, Mahidol University, Nakhon Pathom, Thailand (e-mail: pornchai.bum@mahidol.ac.th).

[PDF]

Cite: Bordin Wanichodom, Nont Neamsuwan, and Pornchai Bumroongsri, "A Neural Network Based Soft Sensor for Online Vapor Product Quality Estimation of a Refinery Debutanizer Column," International Journal of Machine Learning and Computing vol.5, no. 5, pp. 388-391, 2015.

General Information

  • E-ISSN: 2972-368X
  • Abbreviated Title: Int. J. Mach. Learn.
  • Frequency: Quaterly
  • DOI: 10.18178/IJML
  • Editor-in-Chief: Dr. Lin Huang
  • Executive Editor:  Ms. Cherry L. Chen
  • Abstracing/Indexing: Inspec (IET), Google Scholar, Crossref, ProQuest, Electronic Journals LibraryCNKI.
  • E-mail: ijml@ejournal.net


Article Metrics in Dimensions