Accelerated Test and Statistics Model Analysis of Degradation Performance for PV Module Lifetime Prediction

appropriateness of the model describing a product’s degradation path. we propose a relationship between the product’s lifetime and stress variables. The method can be used for highly reliable products. For PV module makers, we hope to have a cost effective way but accuracy method for lifetime prediction. In this study, a general linear degradation model was proposed . The unit-to-unit variation of modules can be considered simultaneously with the time-dependent structure in degradation paths. Based on the proposed degradation model, an implicit expression of a product’s lifetime distribution, and its corresponding mean-time-to-failure (MTTF) was derived. By using the profile likelihood approach, maximum likelihood estimation of parameters, a product’s MTTF can be obtained. In the illustrative example, we use the accelerated degradation data for PV module to address the proposed method. In this study, motivated by the data of PV module power loss, a general linear degradation model which takes “unit-to unit variation”, “time-correlated structure”, and “measurement error” into considerations simultaneously was proposed. The product’s lifetime under the normally used conditions can be obtained. The estimates are reasonable for practical condition from the experiences of engineering and the viewpoint of material property.