Modeling delayed coking plant via RBF neural networks

Author

Zhang, Kejin ; Yu, Jinshou

Author_Institution

Res. Inst. of Autom., East China Univ. of Sci. & Technol, China

Volume

5

fYear

1999

fDate

1999

Firstpage

3483

Abstract

According to the mechanism analysis of a delayed coking plant, combined with the on-site process data, the multiple variables model of liquid products of the delayed coking plant is built by RBF neural networks (RBFNN). This RBFNN-based model provides yield ratio of gasoline, diesel oil, coker gas-oil and general yield ratio of liquid products simultaneously. Results show that the model is practically equivalent and its generalization ability is satisfactory. The simulation results are satisfied and show fine practical value for industrial production operation and optimization control

Keywords

oil refining; optimal control; process control; radial basis function networks; RBF neural networks; coker gas-oil; delayed coking plant; diesel oil; gasoline; generalization ability; industrial production operation; liquid products; multiple variables model; on-site process data; optimization control; yield ratio; Artificial neural networks; Automation; Delay; Electronic mail; Feeds; Furnaces; Neural networks; Petroleum; Recycling; Temperature;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks, 1999. IJCNN '99. International Joint Conference on

Conference_Location

Washington, DC

ISSN

1098-7576

Print_ISBN

0-7803-5529-6

Type

conf

DOI

10.1109/IJCNN.1999.836226

Filename

836226