CONSERVATIVE ESTIMATION OF PROBABILITY OF FAILURE

作者： Victor Picheny , Nam-Ho Kim , Raphael Haftka , Jorg Peters

DOI: 10.2514/6.2006-7038

关键词:

摘要: The paper provides a review of how to estimate probability failure from small sample data, and shows that the usual estimators parameters cumulative distribution function are biased, can lead unconservative estimations. Then, it explores different ways make this estimation conservative: one is based on adding constraints when distributions fitted; second use bootstrap methods. We explore relationship between chance conservative accuracy estimate. In particular, we study case want achieve 95% have estimators. Finally, these methods applied problem composite panel under thermal loading.