Sensor Failure Mode Detection and Self-Validation

Now showing item 1 of 1

Title:	Sensor Failure Mode Detection and Self-Validation
Author:	Abhinav, Abhinav
Description:	The validity, reliability and the performance of complex systems depend on the reliability of the sensors that are used for monitoring and diagnosis of the systems. Hence, it becomes very important to effectively diagnose the failure of sensors. Most of the sensor failure techniques presently in use or proposed by researchers require detail knowledge of the sensor model. First, in this thesis based on the hypothesis that sensor faults generate signature pattern in the output domain, a sensor fault classification is presented. Second, for real-time detection of these failure modes, a sensor and process independent fault detection scheme is developed, which uses threshold checks of different statistical features like limit, mean, and standard deviation for abnormalities. A Matlab tool is also been developed based on the proposed scheme and is tested on various types of sensors for simultaneous detection of the presented failure modes. Third, a detailed literature review on Self-Validating (SEVA) sensor technology is presented and an algorithm for self-validation is developed, using uncertainty as the measure of confidence in the data and the past good value as the validated measurement.
Permanent Link:	http://rave.ohiolink.edu/etdc/view?acc_num=ucin1227254283 http://hdl.handle.net/2374.OX/105642
Date:	2008

Files in this item

Files	Size	Format	View
There are no files associated with this item.

Now showing item 1 of 1