Le 23 janvier 2020

​Optimal maintenance policy for a system subject to shocks and degradation under uncertain parameters

  Firoozeh Haghighi

Abstract:

It is of great importance to avoid a failure of a system when its failure during actual operation is costly and/or dangerous. In such a situation, it is wise to replace the system before failure as a preventive maintenance policy. We studied the problem of optimal maintenance policy for a single-unit system which is subject to shocks and degradation. The shocks could be fatal or effective and arrive in accordance with a counting process. The degradation path follows a non-decreasing arbitrary function and the impact of effective shocks on the degradation of system is considered through the virtual age concept. First, a degradation-threshold-shock model is developed and then, under uncertain model parameters, a rolling horizon algorithm is proposed to obtain a dynamic optimal preventive level of degradation while taking to account the number of receiving shocks and theirs magnitudes, as well as the current damage of system at each pre-specified inspection time. Policy updating is done based on the rolling horizon approach via Bayesian method.

Keywords: Condition-based maintenance, General degradation path model, Long-run cost rate, Random Shocks, Rolling horizon approach, Virtual age

Firoozeh HAGHIGHI is an Associate Professor in the Department of Mathematics, Statistics and Computer Science at University of Tehran, Iran. She received her B.S. and M.S. degrees in Statistics from Shahid Beheshti University, Tehran, Iran, in 1995 and 1998. She also received her PhD degree in Statistics (specialty: Reliability) from University of Bordeaux II, Bordeaux, France in 2004.  Firoozeh HAGHIGHI has published more than 38 papers in journals and conferences proceedings. Her research interests are Accelerated life testing, Degradation models, Competing risks, Maintenance and Quality control.