基于极大似然估计的加速因子计算方法

目的在产品寿命服从参数未知指数分布的情况下,分析定时转换步加试验得到的数据,估算相关的加速因子。方法列出各应力下的寿命分布密度函数形式,通过加速因子建立各分布密度函数之间的联系,以条件概率的方式处理步加试验每一阶段开始时的累积损伤问题,以极大似然估计的方法进行参数估计。结果得出了一种新的估算加速因子的方法,计算获得了产品在所分析应力下加速因子与应力水平之间的关系。结论利用该方法能很好地处理某应力下无失效发生的情况。

Calculation of Acceleration Factor Based on Maximum Likelihood Estimation

Objective To estimate the associated acceleration factor by analyzing the Trial data of step-stress accelerated life tests ( SSALT) with Type-I censored data, assuming the lifetime of products followed exponential distribution. Methods Lifetime distribution density functions under different stresses were listed. Linkages were established between the distribution density functions by acceleration factors ( AF) . And then conditional probability distribution was used to deal with the problem of cumulative damage at the beginning of each stage by stepwise analysis of the testing. The AFs were estimated by maximum likelihood estimation. Results The paper offered a new method to deal with the cumulative damage. The relationship between AFs and stress levels under the tested stress was calculated. Conclusion It is a feasible method to cater the situation that no failure occurs under certain stress.