Method of automated, continuous monitoring to detect faults in complicated dynamic system based on hidden-Markov-model (HMM) approach. Simpler than another, recently proposed HMM method, but retains advantages of that method, including low susceptibility to false alarms, no need for mathematical model of dynamics of system under normal or faulty conditions, and ability to detect subtle changes in characteristics of monitored signals. Examples of systems monitored by use of this method include motors, turbines, and pumps critical in their applications; chemical-processing plants; powerplants; and biomedical systems.
Improved Hidden-Markov-Model Method Of Detecting Faults
NASA Tech Briefs ; 18 , 8
1994-08-01
Miscellaneous
No indication
English
Improved Hidden-Markov-Model Method of Detecting Faults
Online Contents | 1994
Detecting Faults By Use Of Hidden Markov Models
NTRS | 1995
|Detecting Faults by Use of Hidden Markov Models
Online Contents | 1995
Hidden Markov Model-based population synthesis
Online Contents | 2016
|