Redundancy allocation problem of phased-mission system with non-exponential components and mixed redundancy strategy

• Published in: Reliability engineering & system safety Vol. 199; pp. 106903 - 11
• Main Authors: Li, Xiang-Yu, Li, Yan-Feng, Huang, Hong-Zhong
• Format: Journal Article
• Language: English
• Published: Barking Elsevier Ltd 01.07.2020; Elsevier BV
• Subjects: Approximation; Approximation method; Genetic algorithms; Improved genetic algorithm; Markov processes; Mixed redundancy strategy; Optimization; Phased-mission system; Propulsion systems; Redundancy; Reliability engineering; Semi-Markov process; Spacecraft; Subsystems; System reliability; System structure optimization; System structure optimization; Semi-Markov process; Phased-mission system; Improved genetic algorithm; Mixed redundancy strategy
• ISSN: 0951-8320; 1879-0836
• DOI: 10.1016/j.ress.2020.106903

•The mixed redundancy strategy is introduced to the RAP of phased mission system.•The Semi-Markov process is used to evaluate the non-exponential PMS.•An optimization procedure based on the Genetic Algorithm is given.•A spacecraft propulsion system is used to illustrate the proposed method. The redundancy allocation problem (RAP) has been widely studied, which aims to identify the optimal design to achieve the best system indicators, such as system reliability or availability, under certain constraints. This problem has been studied well in the single phased mission systems. But many practical systems, like the manmade satellites or spacecraft, perform a series of tasks in multiple, consecutive, non-overlapping time durations (phases). For these multi-phased systems, dependencies among the phases need to be fully considered. To address this problem, the RAP of the phased mission systems (PMSs) is studied in this paper. Moreover, in order to improve system reliability as much as possible, the mixed redundancy strategy where both active and cold standby components can be simultaneously used in a subsystem is applied. The non-exponential components, which are more practical, are also considered. The Semi-Markov process (SMP) as well as a numerical approximation method are used to deal with the dynamic non-exponential components. Then, an improved Genetic algorithm (GA) is used to determine the types and quantities of active and cold-standby components in each subsystem to optimize the whole system. At last, the propulsion system in a spacecraft is optimized to illustrate the proposed method.