Process fault detection and signal reconstruction using statistical and advanced techniques within the water industry

Author

Fletcher, I. ; Adgar, A. ; Boehme, T.J. ; Cox, C.S.

Author_Institution

Sch. of Eng. & Adv. Technol., Univ. of Sunderland, Sunderland, UK

fYear

1999

fDate

Aug. 31 1999-Sept. 3 1999

Firstpage

2904

Lastpage

2909

Abstract

The efficient and robust operation of any industrial system is critically dependant upon the quality of the measurements made. This paper investigates techniques that are not only capable of identifying faulty sensors, but also of estimating values to permit continued plant operation. These techniques being of particular value when the luxury of sensor redundancy can not be afforded. A comparison between Statistical and Artificial Neural Network methodologies is presented and results of the most successful technique applied to data collected from a Water Treatment Plant are shown.

Keywords

environmental science computing; fault tolerant computing; neural nets; sensors; signal reconstruction; statistical analysis; water treatment; ANN; advanced techniques; artificial neural network methodologies; faulty sensors; plant operation; process fault detection; sensor redundancy; signal reconstruction; statistical techniques; water industry; water treatment plant; Artificial neural networks; Data models; Fault detection; Principal component analysis; Process control; Sensors; Standards; Artificial Neural Networks; Fault detection; Principal Component Analysis; Water Treatment;

fLanguage

English

Publisher

ieee

Conference_Titel

Control Conference (ECC), 1999 European

Conference_Location

Karlsruhe

Print_ISBN

978-3-9524173-5-5

Type

conf

Filename

7099769