Life prediction of Tantalum capacitor based on gray theory optimization model

The reliability level of capacitors significantly affects reliability and maintenance costs of those facilities. For military-grade Tantalum capacitors, it is difficult to assess the reliability by using the traditional time-to-failure analysis method. Based on gray theory, an adaptive double-parameter prediction model, whose parameters were computed by using recognition method, was proposed in this paper. Also the residual test and poster checking were used to check the proposed model. Based on failure thresholds, the life of each sample under its accelerated stress can be predicted with the proposed model. Then distribution test for pseudo-life of each sample was done and the distributed parameters could be estimated. At last the life under normal stress was extrapolated by using Arrehenius model. The results show that the proposed model is superior to GM(1,1). And the predicted life is consistent with the real life.