Reliability Redundancy Allocation Problems (RRAPs) are the optimization problems that involve treating the reliability of components within a subsystem as a decision variable, where all the redundant components in each subsystem are considered homogenous, that is, components in subsystems share the same reliability. However, this assumption is not very suitable for real-world problems. This paper implements an enhanced version of RRAP with nonhomogeneous components that can have different reliability levels in the same subsystem. Such RRAP with component mixing (RRAP-CM) transforms the conventional RRAP model into a more intricate and heterogeneous one. A nonlinear mixed integer mathematical model is formulated and applied to three benchmark test problems. Additionally, the paper develops a robust meta-heuristic algorithm, the Opposition-Based Tunicate Swarm Algorithm (OBTSA), to solve the complex mathematical models of RRAP-CM. To establish the superiority of the proposed work, a detailed comparison is presented with the results of the best designs for benchmark systems in the literature with the maximum possible improvement percentage (MPI%). Also, to validate the adaptability of the work, a real-life example of the water quality monitoring system is taken into consideration and solved for maximum reliability. The outcomes in all systems surpass the previously obtained best results.
Choudhary et al. (Mon,) studied this question.