The reliability of each component in a system plays a crucial role, as any malfunction can significantly reduce the system's overall lifespan. Optimizing the arrangement and sequence of heterogeneous components with varying lifespans is essential for enhancing system stability. This paper addresses the redundancy allocation problem (RAP) by determining the optimal number of components in each subsystem, considering their sequence, and optimizing multiple criteria such as reliability, cost uncertainty, and weight. A novel approach is introduced, incorporating a switching mechanism that accommodates both correct and defective switches. To assess reliability benefits, Markov chains are employed, while cost uncertainty is evaluated using the Monte-Carlo method with risk criteria such as percentile and mean-variance. The problem is solved using a modified genetic algorithm, and the proposed method is benchmarked against alternative approaches in similar scenarios. The results demonstrate a significant improvement in the Model Performance Index (MPI), with the best RAPMC solution under a mixed strategy achieving an MPI of 0.98625, indicating superior model efficiency compared to previous studies. Sensitivity analysis reveals that lower percentiles in the cost evaluations correlate with reduced objective function values and mean-variance, confirming the model's robustness in managing redundancy allocation to optimize reliability and control cost uncertainties effectively.
