为了有效地利用来自退化试验或加速退化试验的信息,提出了一种新的采样策略来获取退化过程超过一系列伪失效阈值的首次通过时间数据,给出了有关模型假设和详细的伪失效时间数据分析步骤,利用极大似然估计获取时变参数分布的参数估计,根据拟合的分布参数函数预测将来的失效时间分布,得到了一种基于伪失效时间数据的可靠性分析方法。最后,利用一个疲劳裂纹增长过程例子说明失效时间分布预测过程并验证提出方法的有效性。
To make use of the information from the degradation test or the accelerated degradation test sufficiently,a new sample strategy was presented to achieve the first passage times of degradation processes to a series of pseudo-failure thresholds.The model assumptions and the concrete analysis steps for pseudo-failure times were introduced.The graphic method was used to analyze these data and the maximum likelihood method was used to estimate the parameters of the time varying distribution.Predictions of the failure time distribution in the future were then implemented from the fitted distribution parameter functions.Hence,a new reliability analysis method based on pseudo-failure times was presented.The methods were illustrated by an example to predict the failure time distribution of the fatigue crack growth processes.