A new model for reliability redundancy allocation problem with component mixing

Reliability Redundancy Allocation Problem (RRAP) is a well-known problem in reliability optimization. In most of the previous research in RRAP, it is assumed that the components allocated to each subsystem are of the same type. This assumption limits the possible choices for the subsystem. In this paper, a new RRAP is proposed where the components allocated to each subsystem can be non-identical, and the redundancy strategy is a decision variable that can be active, standby, mixed, or none. In the proposed RRAP, both the active and the standby components can be independently of different types, which has not been considered in any of the previous studies. Since RRAP is an NP-hard problem, an adapted genetic algorithm is presented to search for the optimal solution of the proposed problem. Finally, some well-known problems are investigated to compare the proposed model with the previous ones. The obtained results show that the proposed RRAP can improve the reliability of the system in different cases by 10 % to 63 % in terms of the maximum possible improvement index.

Automated summary

This paper proposes a new Reliability Redundancy Allocation Problem (RRAP) where non-identical components can be allocated to each subsystem and the redundancy strategy is variable. An improved genetic algorithm is presented to solve the proposed problem. The results demonstrate that the proposed RRAP outperforms the previous models in terms of reliability improvement.