少数の供試品による信頼性評価法の研究
スポンサーリンク
概要
- 論文の詳細を見る
Hypothesis about the reliability of units in a lot, to which the stress-strength model is applied, can be tested by generally sampling both the environment stress Z, and the strength of the unit Y, or directly sampling the attribute of interest Y<Z. When the strength Y depends (linearly) on a explanatory variable X, it is often more efficient to test using the samples that are produced controling the explanatory variable artificially. This study discusses the efficient choice of the control variable X. General guidelines for this choice are suggested, evaluated under some hypothelical test specification and also discussed comparing the compressed limits methods.
- 社団法人日本品質管理学会の論文
- 1990-01-15