可靠性增长试验中一种新的参数估计法

基金项目:

国家基金(9871003)

博士点基金(97000139)

详细信息

A New Process of Estimating Parameter in Reliability Growth Experiment

摘要: 本文在充分利用可靠性增长试验中各阶段试验数据评定现阶段可靠性指标的基础上对成败型和指数寿命型试验下所得的抽样数据，采用一种新的样本空间排序法─字典排序法给出了现阶段可靠性的点估计和置信下限．从理论和实例两个方面论证了它比仅利用现阶段数据的经典方法具有一致的优良性，改进了房祥忠等[4]相应研究中的不足．而文中定理2.4定理3.4的结论不仅与可靠性增长试验的直观背景相吻合，还大大提高了这种新方法的实际可操作性。
- 可靠性增长试验 /
- 点估计 /
- 置信下限
Abstract: Based on the good utility of experimental data of various period in reliability growth experiment to evaluate reliability index in the present case according to the sampling data from binary and exponential life experiment, a new sample space sequencing （i.e. lexicographic order） is given to provide a point estimator and confidence lower bound of the reliability of current period. From theory and instances, it has been proved that this procedure, which overcomes some shortcomings concernedoby Huang [4], is uniformly better than the classical approach. The conculsions of Theorem 2.4 and 3.4 not only include practical background of reliable growth experiment, but also improve greatly the practicable operation on this new process.