Virtual Instruments Based on Stacked Neural Networks to Improve Product Quality Monitoring in a Refinery

Author

Fortuna, L. ; Giannone, P. ; Graziani, S. ; Xibilia, M.G.

Author_Institution

Universita degli Studi di Catania

Volume

3

fYear

2005

fDate

16-19 May 2005

Firstpage

1889

Lastpage

1893

Abstract

A virtual instrument, based on neural networks, for the estimation of octane number in the gasoline produced by refineries is introduced. The stacking approach is proposed to improve the estimation performance of the instrument. The validity of the proposed approach has been verified by comparison with the performance of traditional modeling techniques. The proposed virtual instrument can be used during the maintenance phases of hardware devoted to the measurement of the octane number

Keywords

chemical variables measurement; computerised monitoring; estimation theory; neural nets; petroleum; petroleum industry; virtual instrumentation; gasoline refineries; octane number estimation; product quality monitoring; stacked neural networks; virtual instruments; Delay estimation; Hardware; Industrial plants; Instruments; Intelligent networks; Neural networks; Particle measurements; Petroleum; Refining; Stacking; industrial plants; modeling; neural networks; quality coontrol;

fLanguage

English

Publisher

ieee

Conference_Titel

Instrumentation and Measurement Technology Conference, 2005. IMTC 2005. Proceedings of the IEEE

Conference_Location

Ottawa, Ont.

Print_ISBN

0-7803-8879-8

Type

conf

DOI

10.1109/IMTC.2005.1604500

Filename

1604500