Improving Computational Efficiency of Prediction in Model-based Prognostics Using the Unscented Transform

Matthew Daigle and Kai Goebel
Submission Type: 
Full Paper
AttachmentSizeTimestamp
phmc_10_073.pdf775.66 KBOctober 12, 2010 - 8:34am

Model-based prognostics captures system knowledge in the form of physics-based models of components, and how they fail, in order to obtain accurate predictions of end of life (EOL). EOL is predicted based on the estimated current state distribution of a component and expected profiles of future usage. In general, this requires simulations of the component using the underlying models. In this paper, we develop a simulation-based prediction methodology that achieves computational efficiency by performing only the minimal number of simulations needed in order to accurately approximate the mean and variance of the complete EOL distribution. This is performed through the use of the unscented transform, which predicts the means and covariances of a distribution passed through a nonlinear transformation. In this case, the EOL simulation acts as that nonlinear transformation. In this paper, we review the unscented transform, and describe how this concept is applied to efficient EOL prediction. As a case study, we develop a physics-based model of a solenoid valve, and perform simulation experiments to demonstrate improved computational efficiency without sacrificing prediction accuracy.

Publication Control Number: 
073
Submission Keywords: 
model-based prognostics
particle filters
unscented transform
solenoid valve
Submitted by: 
  
 
 
 

follow us

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