Abstract:
Predictive maintenance of physical systems can only be achieved by monitoring their most critical elements to track their health assessment during operation. The acquired...Show MoreMetadata
Abstract:
Predictive maintenance of physical systems can only be achieved by monitoring their most critical elements to track their health assessment during operation. The acquired data is processed to extract relevant features, which are used to estimate the state of the system at any time and detect any loss of performance that may occur due to the critical element. We propose in this work an architecture of generic method to supervise this critical element and generate a Health Indicator (HI) for the physical system. The generated HI takes into account the evolution in time of the healthy status of the physical systems. The proposed method is based on sensors data that allow us to extract in real time the values of features constituting themselves the HI construction bloc input, through several HI obtaining test. Block diagram of the approach is made, then checked using benchmark data taken from “NASA data repository prognosis” associated to an element used in different operating conditions are checked. This approach is classified as data driven method which use sensors data that inform us about the real-time values of features.
Date of Conference: 24-26 October 2016
Date Added to IEEE Xplore: 05 January 2017
ISBN Information:
Electronic ISSN: 2327-1884