摘要
在综合双侧规格限下研究方差未知的正态分布产品可靠性抽样检验方法 .检验统计量取为不可靠度 p的极大似然估计 p.当 p小时 ,p的分布基本上仅依赖于 p ,对 p分解为下侧不可靠度和上侧不可靠度的依赖轻微 .因此 ,利用单侧规格限下的抽样检验方案近似地导出综合双侧规格限下的抽样检验方案 .随机模拟结果表明 ,给出的近似方法是有效的 ,抽样检验方案是合理的 .此外 ,检验统计量
This paper studies reliability sampling inspection method for items whose quality characteristic has a normal distribution with an unknown variance under double combined specification limits. The test statistic is taken to be the maximum likelihood estimator (MLE), \$\$, of the unreliability \$p\$. When \$p\$ is small, the distribution of \$\$ is essentially dependent only on \$p\$ and only very slightly on the partition of \$p\$ into its components, the lower unreliability and the upper unreliability. The sampling inspection plans under the single specification limit are used to approximately derive the sampling inspection plans under the double combined specification limits. Simulation results show that the approximation method is efficient and the derived sampling inspection plans are reasonable. Furthermore, the expression of the test statistic \$\$ is simpler.
出处
《系统工程理论与实践》
EI
CSCD
北大核心
2002年第9期63-69,96,共8页
Systems Engineering-Theory & Practice
基金
国家自然科学基金 (1 9871 0 88)
