Title :
On-line sensor validation of single sensors using artificial neural networks
Author :
Himmelblau, David M. ; Bhalodia, Mohan
Author_Institution :
Dept. of Chem. Eng., Texas Univ., Austin, TX, USA
Abstract :
We examine how the signal from a single sensor from a dynamic process might be analyzed to ascertain whether the sensor signal is valid or not. It is quite possible for the signal to indicate a change has occurred in the process when the sensor itself is what has changed. Two possible approaches to sensor validation discussed here are (1) use of second and higher order statistics rather than the mean of a signal, and (2) modeling the sensor signal via artificial neural networks
Keywords :
calibration; higher order statistics; neural nets; sensors; signal processing; artificial neural networks; high-order statistics; online sensor validation; sensor signal modeling; Artificial neural networks; Chemical sensors; Higher order statistics; Sensor phenomena and characterization; Sensor systems; Sequential analysis; Signal analysis; Signal processing; Statistical analysis; Temperature sensors;
Conference_Titel :
American Control Conference, Proceedings of the 1995
Conference_Location :
Seattle, WA
Print_ISBN :
0-7803-2445-5
DOI :
10.1109/ACC.1995.529354
