Statistical inference of stress-strength reliability of some coherent systems with cold standby redundancy

In this dissertation, statistical inference methods are considered for the series and parallel system structures that comprise original components and standby redundancy components. Estimation of stress-strength reliability and mean remaining strength for the series and parallel systems are obtained when the underlying distributions are exponential. For this purpose, the classical and Bayesian techniques are implemented. Next, corresponding asymptotic confidence intervals via Fisher information matrix or observed information matrix and highest probability density credible intervals are derived. A simulation study is implemented for statistical inference for each system structure. The performances of maximum likelihood and Bayesian estimates are compared via Monte Carlo simulations. As a final step, real-life data are analyzed for each case.