Multi-objective availability and cost optimization by PSO and COA for series-parallel systems with subsystems failure dependencies

Mohamed Arezki Mellal (1) , Enrico Zio (2, 3)

1 UMBB - Université M'Hamed Bougara Boumerdes
2 POLIMI - Politecnico di Milano [Milan]
3 CRC - Centre de recherche sur les Risques et les Crises

System availability and cost are two of the elements of system dependability. Most systems involve subsystems with failure dependencies. The failure dependencies complicate the optimization of those elements. In this paper, the multi-objective optimization problem of availability and cost is addressed for series-parallel systems with subsystems failure dependencies in case of strong dependency. The problem is solved by applying the particle swarm optimization (PSO) algorithm and the cuckoo optimization algorithm (COA). The multi-objective optimization problem is converted to a single one using the weighted-sum method. The results of a system consisting of six subsystems are analyzed with a comparison of the methods. The best value of the system availability and cost, number of function evaluations, CPU time, and standard deviation reveal that the COA has outperformed the PSO.

Mots clés: system availability; system cost; series-parallel system; failure dependencies; weighted-sum method; particle swarm optimization; cuckoo optimization algorithm

Domaines: Sciences de l'ingénieur [physics]

Mohamed Arezki Mellal, Enrico Zio. Multi-objective availability and cost optimization by PSO and COA for series-parallel systems with subsystems failure dependencies. Microprocessors and Microsystems: Embedded Hardware Design , 2022, 89, pp.104422. ⟨10.1016/j.micpro.2021.104422⟩. ⟨hal-03907481⟩