Bayesian analysis of two-phase degradation data based on change-point Wiener process

In degradation test of some products such as plasma display panels (PDPs) and organic light emitting diodes (OLEDs), observed degradation paths tend to exhibit two-phase patterns over testing period. In this paper, we propose a change-point Wiener process (CPWP) model to fit the degradation paths with two-phase pattern mainly in a Bayesian framework. Considering the distinct degradation behaviors between testing units, we assume that degradation rates and change-points vary from unit to unit. Then hierarchical Bayesian approach is employed to estimate the parameters in the CPWP model. For comparison purpose, we also develop the maximum likelihood (ML) method. The results from simulation study show that the hierarchical Bayesian approach provides more robust inference on the model parameters than ML method. The analysis of OLED degradation data presents that the CPWP model outperforms three other existing models in terms of reliability prediction.