Reliability Allocation Problem in Series-Parallel Systems

Alice Yalaoui (1) , Farah Belmecheri (1) , Eric Chatelet (2) , Farouk Yalaoui (1)

1 LOSI - Laboratoire d'Optimisation des Systèmes Industriels
2 LM2S - Laboratoire Modélisation et Sûreté des Systèmes

Reliability optimization is an important step in industrial systems design. In order to develop a reliable system, designers may introduce different redundant technologies with the same functionality in parallel. In this paper, each technology is assumed to be composed of series components. The obtained configuration belongs to the series-parallel systems. The presented tool is for the design or the improvement of such systems, in order to minimize the system cost with a reliability constraint. The aim is to find the reliability to allocate to each component in order to minimize the total cost, such that the global system reliability verifies a minimal level constraint. This problem is known to be NP-hard. In this paper, a metaheuristic approach, based on the Ant Colony Optimization technics (ACO), is used in order to improve an existing approach. The experimental results, based on randomly generated instances, outperform the one of previous method dedicated to this problem.

Domaines: Recherche opérationnelle [math.OC]

Alice Yalaoui, Farah Belmecheri, Eric Chatelet, Farouk Yalaoui. Reliability Allocation Problem in Series-Parallel Systems. International Journal of Applied Evolutionary Computation, 2011, 2 (1), pp.1-17. ⟨10.4018/jaec.2011010101⟩. ⟨hal-02562554⟩