Lifetime Inference of Skew-Wiener Linear Degradation Models
Degradation models are widely used to assess the lifetime information for highly reliable products which have quality characteristics whose degradation over time can be related to reliability. The performance of a degradation model depends strongly on the appropriateness of the model for describing a products degradation path. Conventionally, the random (or mixed) effect model is one of the bestknown approaches in the literature. To model unit-tounit variability, normal distribution is commonly adopted to model the random effect in the degradation model. However, this assumption may not be so appropriate in practical applications. In this paper, motivated by laser data, this distribution is relaxed by a skew-normal distribution which provides flexibility in capturing a broad range of non-normal and asymmetric behavior in unit-to-unit variation. Based on the proposed degradation model, we first derive analytical expressions for the products lifetime distribution and its corresponding meantime-to-failure (MTTF). Finally, the laser degradation data is used to illustrate the proposed model.
Degradation model EM algorithm Wiener process Skew-normal First passage time
CHIEN-YU PENG SHENG-TSAING TSENG
Institute of Statistical Science, Academia Sinica, Taipei, 11529, Taiwan Institute of Statistics, National Tsing-Hua University, Hsinchu, 30013, Taiwan
国际会议
北京
英文
428-434
2011-06-20（万方平台首次上网日期，不代表论文的发表时间）