Abstract
This paper deals with a parallel load-sharing reliability system with cold standby redundancy and ample repair facilities. That is, we have n identical parallel units, of which at most k units are operating simultaneously. If less than k units are available, the system operates at a proportionally reduced level. For this system, an approximate method is given for the calculation of the probability distribution of that proportion of the system capacity that cannot be used in a given time period. The method is based on an approximation of the k-out-of-n multistate system by a two-state single component. Validation of the approximation using Monte-Carlo simulation shows satisfactory performance. Also, sensitivity results are given, showing in particular a decreasing sensitivity of the measures of performance to the distributional form of the unit lifetimes and repair times as the size of the system increases. Furthermore, it is found that the effect of the distributional form of the unit lifetimes dominates that of the unit repair times
Original language | English |
---|---|
Pages (from-to) | 66-77 |
Journal | European journal of operational research |
Volume | 36 |
Issue number | 1 |
DOIs | |
Publication status | Published - 1988 |
Keywords
- Markov process
- (Un)effectiveness
- k-out-of-n system
- Production
- Phase-type distribution