A maintenance policy for a system with multi-state components: an approximate solution

Abstract

Cataloged from PDF version of article.For maintenance and quality assessment purposes, various performance levels for both systems and components are identified, usually as a function of the deterioration. In this study, we consider a multicomponent system where the lifetime of each component is described by several stages, (0,…,S), which are further classified as good, doubtful, preventive maintenance due (PM due) and down. A control policy is suggested where the system is replaced when a component enters a PM due or a down state and the number of components in the doubtful states (K,…,S−2) is at least N. All maintenance activities are assumed to take negligible time. The exact description of the underlying stochastic model under the policy is very complicated. We therefore propose some approximations, which allow an explicit expression for the long run average cost function, which is minimized w.r.t. (K,N) by numerical methods. Sensitivity of the model to system parameters and the performance of the approximation are investigated through several examples

On problems of multicomponent system maintenance modelling

“…Gurler and Kaya [31] considered a multicomponent system where the lifetime of each component is described by several stages, which are further classified as good, doubtful, PM due, and down. The system is composed of n identical and independently operating components that are connected in series.…”

Approximate methods for optimal replacement, maintenance, and inspection policies

“…However, to suppose the unit fails with constant probability p is too stronger to be applicable, even though this policy has been used for Weibull and gamma distribution cases [8,17]. Other works, such as an approximate solution of a maintenance policy for a system with multi-state components [4], an approximate inspection interval for production processes with finite run length [3], and approximations to determine the optimal replacement times of a sequential age replacement policy for a finite time horizon [6], have been discussed.…”

Improved Acquisition for System Sustainment: Multiobjective Tradeoff Analysis for Condition-Based Decision-Making

“…Previous Markov chain-based models were limited to few components. An opportunistic maintenance policy was modeled by Gürler and Kaya (2002) and van der Duyn Schouten and Vanneste (1993) for identical multi^Åèìáëáíáçå=oÉëÉ~êÅÜ=mêçÖê~ã= dê~Çì~íÉ=pÅÜççä=çÑ=_ìëáåÉëë=C=mìÄäáÅ=mçäáÅó -27 -k~î~ä=mçëíÖê~Çì~íÉ=pÅÜççä= component system. Sheu and Jhang (1996) modeled a similar kind of problem with a two-stage opportunistic policy.…”