Abstract: Based on Arrhenius Power law model, the numerical methods for analysed grouping the data of Constant Stress Accelerated Reliability Growth Testing(CSARGT) were given.They included the statistical tests of four fundamental hypotheses of CSARGT, the Maximum Likelihood Estimation(MLE) and interval estimation of the parameters of the power law model and MTBF at accelerated stress levels S i ,the Best Linear Unbiased Estimation(BLUE) of the coefficient of acceleration equation, the interval estimations of acceleration factor and converted MTBF at normal stress level S o. These methods were illustrated.

Key words: Reliability growth testing;Acceleration life test;Fault analysis;Grouped data;Numerical method and Procedure;Acceleration factor

摘要： 基于Arrhenius 幂律模型 ,给出了分析恒定应力加速可靠性增长试验 (CSARGT)分组数据的数值方法 ,它们包括对CSARGT四项基本假定的统计检验 ,在诸加速应力水平Si 下幂律模型参数与MTBF的极大似然估计(MLE)与区间估计 ,加速方程系数的最佳线性无偏估计 (BLUE) ,加速系数与正常应力水平S0 下折算的MTBF的区间估计 ,并用数值例说明了这些方法。

关键词: 可靠性增长试验;加速寿命试验;故障分析;分组数据;数值法和计算方案;加速系数