Failure Prognostics with Missing Data Using Extended Kalman Filter

Wlamir O. L. Vianna and Takashi Yoneyama
Submission Type: 
Full Paper
AttachmentSizeTimestamp
phmc_16_004.pdf933.43 KBAugust 5, 2016 - 8:01am

Failure prognostics can provide benefits in operation and maintenance of equipments by predicting when the component is going to fail and consequently acting at the most appropriate time. In several situations degradation estimations are sparse or missing estimations are present at collected data. Considering these situations, a failure prognostics method was proposed considering the usage of the extended version of the Kalman filter. This method was submitted to analysis considering some hydraulic levels data collected from four different aircrafts. In this study a prognostic model was estimated by the filter and then future values of hydraulic level as well as the remaining useful life interval were obtained considering a set of Monte Carlo simulations and a failure probability distribution approximation. Results evidenced the benefit of this method to properly prognose the system.

Publication Year: 
2016
Publication Volume: 
7
Publication Control Number: 
004
Page Count: 
5
Submission Keywords: 
failure prognostics
extended Kalman filter
Missing Data
Submission Topic Areas: 
Data-driven methods for fault detection, diagnosis, and prognosis
  
 
 
 

follow us

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