Reliability prediction method with field environment variation

This paper presents a novel method of field reliability prediction considering environment variation and product individual dispersion. Wiener diffusion process with drift was used for degradation modeling and a link function which presents degradation rate is introduced to model the impact of varied environment and individual dispersion. Gamma, transformed-Gamma (T-Gamma) and Normal distribution with different parameters are employed to model right-skewed, left-skewed and symmetric stress distribution in the study case. Results show obvious difference in reliability, failure intensity and failure rate compared to constant stress situation and each other. It indicates that properly modeled (proper distribution type and parameters) environmental stress is the fundamental of varied environment oriented reliability prediction. In a linear drift degradation process and Arrhenius-typ e link function situation, it is reasonable not concerning about product individual dispersion because the impact is barely small, while other situations can be studied in the same way proposed in this present paper.