Abstract
In this paper, we consider optimum constant-stress accelerated life testing (ALT) plans with multiple estimation objectives under log-normal distribution. We propose two kinds of design criterion: one is minimizing the weighted sum of the absolute difference between the variance of each objective estimator and its minimum variance, and the other is minimizing the weighted sum of the relative difference. The plans based on the two criteria are compared with testing plans for single estimation objective, and D-optimal testing plans. Comparison results show that the proposed plans provide more accurate estimation than D-optimal plans do, and is more robust than the testing plans for single estimation objective.
| Original language | English |
|---|---|
| Pages | 6-9 |
| Number of pages | 4 |
| State | Published - 2009 |
| Event | 15th ISSAT International Conference on Reliability and Quality in Design - San Francisco, CA, United States Duration: 6 Aug 2009 → 8 Aug 2009 |
Conference
| Conference | 15th ISSAT International Conference on Reliability and Quality in Design |
|---|---|
| Country/Territory | United States |
| City | San Francisco, CA |
| Period | 6/08/09 → 8/08/09 |
Keywords
- Constant-Stress ALT
- Log-Normal Distribution
- Multiple Estimation Objectives
- Optimum Design