Inferential Statistics on Load Sharing System Model with Bathtub Shape Failure Rate under Type II Censoring Scheme

In load-sharing systems, the components are subjected to share the system workload and if any of the components within the system fails, its load is immediately transfer to the rest of the surviving components, due to which the system working efficiency remains unaffected. The present study addresses the problem of parameter estimation of kcomponent load-sharing redundant system model using components' failure data under Type-II censoring scheme. The underlying failure rate of each component is assumed to have bathtub shape. In classical setup, we obtain maximum likelihood estimates and asymptotic confidence intervals for the load-sharing parameters, whereas, Bayesian estimates are obtained under symmetric and asymmetric loss functions assuming joint Dirichlet prior distribution of the unknown parameters. A simulation study is performed for numerical comparison of the different estimation methods.