Title :
Sensor noise fault detection
Author_Institution :
Inst. for Sci. Res., Fairmont, WV, USA
Abstract :
Current sensor FDIR (fault detection, isolation, & recovery) generally focuses on sensor bias and drift anomalies, which require models. However, dead sensors and excessive noise faults are more common in practice. The latter two faults are interesting in that they can be detected using only the measurements from each sensor. The objective of this paper is to show a few ways to detect common sensor faults and thus, enhance sensor reliability by using instrument signals to better advantage.
Keywords :
computerised monitoring; fault diagnosis; noise; reliability; sensors; common sensor fault; dead sensor; drift anomaly; excessive noise fault; fault detection; fault isolation; fault recovery; instrument signal; sensor FDIR; sensor bias; sensor noise fault detection; sensor reliability; statistical fault detection; Analytical models; Bandwidth; Event detection; Failure analysis; Fault detection; Fault trees; Instruments; Redundancy; Signal to noise ratio; Wearable sensors;
Conference_Titel :
American Control Conference, 2003. Proceedings of the 2003
Print_ISBN :
0-7803-7896-2
DOI :
10.1109/ACC.2003.1240506