Optimal Design for Step-Stress Accelerated Degradation Tests
Authors:
- Sheng-Tsaing TsengSTSheng-Tsaing Tseng
- 6 March 2006
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Reliability
- Vol. 55 (1) , 59-66
Abstract
Today, many products are designed to function for a long period of time before they fail. For such highly-reliable products, collecting accelerated degradation test (ADT) data can provide useful reliability information. However, it usually requires a moderate sample size to implement an ADT. Hence, ADT is not applicable for assessing the lifetime distribution of a newly developed or very expensive product which only has a few available test units on hand. Recently, a step-stress ADT (SSADT) has been suggested in the literature to overcome the above difficulty. However, in designing an efficient SSADT experiment, the issue about how to choose the optimal settings of variables was not discussed, such as sample size, measurement frequency, and termination time. In this study, we first use a stochastic diffusion process to model a typical SSADT problem. Next, under the constraint that the total experimental cost does not exceed a predetermined budget, the optimal settings of these variables are obtained by minimizing the asymptotic variance of the estimated 100p/sup th/ percentile of the product's lifetime distribution. Finally, an example is used to illustrate the proposed method.Keywords
This publication has 11 references indexed in Scilit:
- Wiley Series in Probability and Statistics2008
- Determination of burn‐in parameters and residual life for highly reliable productsNaval Research Logistics (NRL), 2002
- Step-Stress Accelerated Degradation Analysis for Highly Reliable ProductsJournal of Quality Technology, 2000
- Designing a degradation experimentNaval Research Logistics (NRL), 1999
- Accelerated Degradation Tests: Modeling and AnalysisTechnometrics, 1998
- A termination rule for degradation experimentsIEEE Transactions on Reliability, 1997
- Experimental Design for a Class of Accelerated Degradation TestsTechnometrics, 1994
- Experimental Design for a Class of Accelerated Degradation TestsTechnometrics, 1994
- Models for Variable-Stress Accelerated Life Testing Experiments Based on Wiener Processes and the Inverse Gaussian DistributionTechnometrics, 1992
- Reliability assessment based on accelerated degradation: a case studyIEEE Transactions on Reliability, 1991