Paper Info
Title
Supervision and prognosis architecture based on dynamical classification method for the predictive maintenance of dynamical evolving systems.
Abstract
In this paper, we are concerned by the improvement of the safety, availability and reliability of dynamical systems’ components subjected to slow degradations (slow drifts). We propose an architecture for efficient Predictive Maintenance (PM) according to the real time estimate of the future state of the components. The architecture is built on supervision and prognosis tools. The prognosis method is based on an appropriated supervision technique that consists in drift tracking of the dynamical systems using AUDyC (AUto-adaptive and Dynamical Clustering), that is an auto-adaptive dynamical classifier. Thus, due to the complexity and the dynamical of the considered systems, the Failure Mode Effect and Criticity Analysis (FMECA) is used to identify the key components of the systems. A component is defined as an element of the system that can be impacted by only one failure. A failure of a key component causes a long downtime of the system. From the FMECA, a Fault Tree Analysis (FTA) of the system are built to determine the propagation laws of a failure on the system by using a deductive method. The proposed architecture is implemented for the PM of a thermoregulator. The application on this real system highlights the interests and the performances of the proposed architecture.
Year
DOI
Venue
2015
10.1016/j.ress.2014.12.005
Reliability Engineering & System Safety
Keywords
DocType
Volume
Non-stationary environment,Supervision,Prognosis,Predictive maintenance,Evolving systems,Dynamical classification
Journal
136
ISSN
Citations 
PageRank 
0951-8320
0
0.34
References 
Authors
0
3
Name
Order
Citations
PageRank
Moussa Traore100.34
Antoine Chammas200.34
Eric Duviella32311.69