摘要
针对传统指标检验方法进行验证决策所需样本量大的问题,提出一种序贯验后加权检验方法(se-quential posterior odd test,SPOT),给出系统寿命服从指数分布条件下,应用SPOT方法进行可靠性指标验证时决策阈值的计算模型及给定样本容量时的截尾SPOT方法。由于该方法充分利用了验前获得的关于总体分布未知参数的信息,因而减小了对现场试验子样的需求。最后给出了应用该方法的具体步骤,通过仿真实验证明了该方法的有效性。
To avoid large sample size required by traditional index verification test, a new sequential posterior odd test (SPOT) method is introduced. The algorithm to determine threshold in SPOT method for reliability index verification is presented and the censored SPOT method with given sample size is also discussed with exponential life distribution. Because this method can make full use of prior information about unknown parameter, it can reduce the demand for field testing sample size. Finally, the definite steps of this method are addressed and the simulation example shows that the proposed method is effective.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2006年第8期1282-1284,共3页
Systems Engineering and Electronics
基金
国家自然科学基金资助课题(70571083)
关键词
可靠性
指数分布
SPOT方法
指标验证
贝叶斯分析
reliability
exponential distribution
SPOT method
index verification
Bayesian analysis