A simulation-based optimization approach for free distributed repairable multi-state availability-redundancy allocation problems

• Published in: Reliability engineering & system safety Vol. 157; pp. 177 - 191
• Main Authors: Attar, Ahmad, Raissi, Sadigh, Khalili-Damghani, Kaveh
• Format: Journal Article
• Language: English
• Published: Barking Elsevier Ltd 01.01.2017; Elsevier BV
• Subjects: Algorithms; Availability; Availability-redundancy allocation; Classification; Computer simulation; Distribution-free failure/repair time; Evolutionary algorithms; Genetic algorithms; Mathematical models; Multi-state systems; Multiple objective analysis; Non-dominated sorting genetic algorithm; Optimization; Redundancy; Reliability engineering; Repairable systems; Simulation based optimization (SBO); Sorting algorithms; Availability-redundancy allocation; Repairable systems; Multi-state systems; Simulation based optimization (SBO); Non-dominated sorting genetic algorithm; Distribution-free failure/repair time
• ISSN: 0951-8320; 1879-0836
• DOI: 10.1016/j.ress.2016.09.006

A simulation-based optimization (SBO) method is proposed to handle multi-objective joint availability-redundancy allocation problem (JARAP). Here, there is no emphasis on probability distributions of time to failures and repair times for multi-state multi-component series-parallel configuration under active, cold and hot standby strategies. Under such conditions, estimation of availability is not a trivial task. First, an efficient computer simulation model is proposed to estimate the availability of the aforementioned system. Then, the estimated availability values are used in a repetitive manner as parameter of a two-objective joint availability-redundancy allocation optimization model through SBO mechanism. The optimization model is then solved using two well-known multi-objective evolutionary computation algorithms, i.e., non-dominated sorting genetic algorithm (NSGA-II), and Strength Pareto Evolutionary Algorithm (SPEA2). The proposed SBO approach is tested using non-exponential numerical example with multi-state repairable components. The results are presented and discussed through different demand scenarios under cold and hot standby strategies. Furthermore, performance of NSGA-II and SPEA2 are statistically compared regarding multi-objective accuracy, and diversity metrics. [Display omitted] •A Simulation-Based Optimization (SBO) procedure is introduced for JARAP.•The proposed SBO works for any given failure and repair times.•An efficient simulation procedure is developed to estimate availability.•Customized NSGA-II and SPEA2 are proposed to solve the bi-objective JARAP.•Statistical analysis is employed to test the performance of optimization methods.