D-optimal Design for Accelerated Life Tests for Lognormal Life Distributions

DOI Web Site 9 References

Bibliographic Information

Other Title
  • 対数正規分布に基づく加速寿命試験のD最適計画
  • タイスウ セイキ ブンプ ニ モトズク カソク ジュミョウ シケン ノ D サイテキ ケイカク

Search this article

Abstract

Product lifetime is often evaluated by performing accelerated life testing under conditions more stressful than actual usage conditions. The lifetime is estimated using an acceleration model. The use of a valid model enables efficient testing using a limited number of samples and a limited testing time. We used a log-normal acceleration model with two factors (temperature and relative humidity) to derive a test plan for maximizing estimation accuracy under the constraints of a given testing time and a given number of samples. Use of this model revealed that a D-optimal design is determined by the ranges of two levels of both factors and the failure probability, and that there is a trade-off relationship between these two factors. If the failure probability is larger than 0.8, the equal number of sample allocation has almost the same precision as that of the optimal design. Also the effect on accuracy is examined when pre-set estimates of the parameter values used in the life testing design are different from the values of true parameters.

Journal

References(9)*help

See more

Related Projects

See more

Keywords

Details 詳細情報について

Report a problem

Back to top