Diagnosing Intermittent and Persistent Faults using Static Bayesian Networks

Brian Ricks and Ole Mengshoel
Submission Type: 
Full Paper
Supporting Agencies (optional): 
CMU SV
AttachmentSizeTimestamp
phmc_10_104.pdf163.42 KBSeptember 21, 2010 - 2:14pm

Reliable health management systems are important to NASA in many applications. It is important for any diagnostic algorithm to be able to quickly and and accurately diagnose not only persistent faults, but to distinguish intermittent faults from their persistent counterparts. We introduce in this paper intermittent fault handling techniques employed by the probabilistic diagnostic algorithm ProDiagnose. We describe the intermittent fault handling algorithm used by ProDiagnose and the implementations involved in the probabilistic models. We show by experimentation on an Electrical Power System based on the ADAPT testbed, used in the Diagnostic Challenge Competition (DXC 10), that ProDiagnose can catch intermittent faults quickly and with very high accuracy.

Publication Control Number: 
104
Submitted by: 
  
 
 
 

follow us

PHM Society on Facebook Follow PHM Society on Twitter PHM Society on LinkedIn PHM Society RSS News Feed