A note on a reliability redundancy allocation problem using a tuned parameter genetic algorithm

作     者：Chambari, Amirhossain Sadeghi, Javad Bakhtiari, Fakhri Jahangard, Reza 

作者机构：Islamic Azad Univ Qazvin Branch Young Researchers & Elite Club Qazvin Iran SUNY Binghamton Syst Sci & Ind Engn Binghamton NY 13902 USA Payame Noor Univ West Tehran Branch Dept Business Adm Fac Management Tehran Iran Islamic Azad Univ Dept Business Adm Persian Gulf Int Branch Khorramshahr Iran 

出 版 物：《OPSEARCH》 (运筹学；印度运筹学会志)

年 卷 期：2016年第53卷第2期

页      面：426-442页

学科分类：1201[管理学-管理科学与工程(可授管理学、工学学位)] 07[理学] 070104[理学-应用数学] 0701[理学-数学] 

主　　题：Redundancy allocation problem Series-parallel systems Continuous genetic algorithm Parameter tuning Response surface methodology (RSM) 

摘      要：This paper presents an improved continuous genetic algorithm (CGA) to optimize the reliability redundancy allocation problem (RRAP) which determines the best redundancy strategies, the number of components, and levels of each subsystem to maximize the system reliability. In this system, both active and cold-standby redundancies can be chosen for individual subsystems. The RRAP belongs to NP-hard problems in the computational complexity theory that is the main reason for employing CGA to solve it. In addition, the response surface methodology (RSM) is used to increase the performance of CGA considering the design of experiments. This algorithm employs a new chromosome so that frees offspring to repair during the evolution process. Considering several numerical examples, the proposed algorithm presents better solutions than the previous studies based on the system reliability. Finally, the conclusion and future research are considered.