An investigation on imperialist competitive algorithm for solving reliability-redundancy allocation problems

Reliability-redundancy allocation problems (RRAPs) are optimization models that try to find the optimal number of redundant components and their reliability levels simultaneously. Many studies have been developed to solve RRAPs in recent years. There are some specific RRAP models for various system...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2015 IEEE International Conference on Industrial Engineering and Engineering Management (IEEM) s. 1041 - 1045
Hlavní autoři: Khorshidi, H. A., Gunawan, I., Sutrisno, A., Nikfalazar, S.
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 01.12.2015
Témata:
Analysis of variance
Bridges
Comparison
Imperialist competitive algorithm
Linear programming
Optimization
Reliability engineering
Reliability-redundancy allocation problem
Resource management
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Reliability-redundancy allocation problems (RRAPs) are optimization models that try to find the optimal number of redundant components and their reliability levels simultaneously. Many studies have been developed to solve RRAPs in recent years. There are some specific RRAP models for various system structures to maximize system reliability subject to cost, volume and weight constraints. Different meta-heuristic algorithms have been used in order to reach the best objective function value. In this study, an investigation is done on imperialist competitive algorithm (ICA) to maximize models for series and bridge systems. ICA is used by adjusting different values to algorithm's parameters. This investigation recognizes which combination is the most suitable for solving the RRAPs by ICA. Each combination has been run for 35 times. Therefore, the combinations are compared by descriptive statistics' measures and analysis of variance (ANOVA). Furthermore, the best obtained solution is compared with the previous studies.
DOI:10.1109/IEEM.2015.7385807