Optimal burn-in for minimally repaired components from n subpopulations

Abstract

Burn-in is a widely used technique to screen out defective components before they are put into field operation. In this paper, we propose a burn-in policy for the components from a heterogeneous population composed of [Formula: see text] ordered subpopulations. Assume that components can be minimally repaired when the failure occurs. We consider the total number of minimal repairs observed during burn-in process as a screening rule to remove defective components with poor reliability performance. First, we carry out the probabilistic analysis of the screening policy to show the effectiveness of burn-in. Second, we obtain the optimal burn-in settings which minimize the mean total cost or maximize the probability of passing the mission period. Finally, we give an example to illustrate the proposed model.