Availability Optimization of Parallel-Series System by Evolutionary Computation

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

1 UMBB - Université M'Hamed Bougara Boumerdes
2 LGI - Laboratoire Génie Industriel - EA 2606

This paper addresses the problem of availability optimization of a parallel-series system by using evolutionary techniques. Five techniques are considered, namely the cuckoo optimization algorithm, particle swarm optimization, flower pollination algorithm, differential evolution, and genetic algorithms. The integer values and constraints of cost are handled. The results regarding a system with ten subsystems are compared. It demonstrates that the cuckoo optimization algorithm performs best.

Mots clés: system availability; system cost; parallel-series system; optimization; evolutionary computation

Domaines: Informatique [cs]

Mohamed Arezki Mellal, Enrico Zio. Availability Optimization of Parallel-Series System by Evolutionary Computation. 2018 3rd International Conference on System Reliability and Safety (ICSRS), Nov 2018, Barcelona, Spain. pp.198-202, ⟨10.1109/ICSRS.2018.8688722⟩. ⟨hal-02428549⟩