The robust redundancy allocation problem of series-parallel systems

Redundancy allocation is a technique that has been widely used to improve systems reliability. In this paper, the interval analysis is introduced to represent imprecise component reliabilities, and an order relation is applied in the comparison of interval values. We propose a mathematical model to deal with constrained redundancy optimization of series-parallel systems with interval-valued component reliabilities. The objective function is extended to an interval-valued function in the proposed model, and a dynamic penalty genetic algorithm (GA) is developed to search the best solution and the optimum system reliability. Several numerical examples are given to illustrate the proposed method, and the experimental results have shown that the interval mathematics is an efficient tool to solve the redundancy allocation problem (RAP) of systems with interval-valued component reliabilities. © 2013 IEEE.