Neural networks in process fault diagnosis

Author

Sorsa, Timo ; Koivo, Heikki N. ; Koivisto, Hannu

Author_Institution

Dept. of Electr. Eng., Tampere Univ. of Technol., Finland

Volume

21

Issue

4

fYear

1991

Firstpage

815

Lastpage

825

Abstract

Fault detection and diagnosis is an important problem in process automation. Both model-based methods and expert systems have been suggested to solve the problem, along with the pattern recognition approach. A number of possible neural network architectures for fault diagnosis are studied. The multilayer perceptron network with a hyperbolic tangent as the nonlinear element seems best suited for the task. As a test case, a realistic heat exchanger-continuous stirred tank reactor system is studied. The system has 14 noisy measurements and 10 faults. The proposed neural network was able to learn the faults in under 3000 training cycles and then to detect and classify the faults correctly. Principal component analysis is used to illustrate the fault diagnosis problem in question

Keywords

chemical reactions; computerised pattern recognition; fault location; neural nets; process computer control; expert systems; heat exchanger-continuous stirred tank reactor system; hyperbolic tangent; model-based methods; multilayer perceptron network; neural network architectures; noisy measurements; pattern recognition approach; process automation; process fault diagnosis; training cycles; Automation; Diagnostic expert systems; Fault detection; Fault diagnosis; Inductors; Multilayer perceptrons; Neural networks; Pattern recognition; Principal component analysis; System testing;

fLanguage

English

Journal_Title

Systems, Man and Cybernetics, IEEE Transactions on

Publisher

ieee

ISSN

0018-9472

Type

jour

DOI

10.1109/21.108299

Filename

108299