基于双截尾数据消除法的自适应加速寿命试验数据统计分析

摘要

加速寿命试验数据处理必须将不同应力水平下的失效数据折算累加到同一标准下的时间序列,这一过程复杂繁琐,增加了计算工作量,降低了统计效率,而指数分布的双截尾数据消除法能有效克服这一缺陷.提出了一种基于模型转换的双截尾数据消除法,将数控系统故障间隔时间服从的威布尔分布转化为指数分布,再用双截尾数数据消除法进行计算,进行自适应加速寿命试验数据的统计分析,有效提高了统计分析效率.实际算例验证了方法的可行性.%According to the accelerated life testing data statistics analysis, the different stress level failure data must be converted into the same standard time series. The calculation is complicated and troublesome and the statistical efficiency is low. This defect can be effectively overcome by the double truncation data elimination method of the exponential distribution. So a double truncation data elimination method based on model transformation was proposed. At frist, the Weibull distribution was converted into exponential distribution because the numerical control system lifetime obeys the Weibull distribution, then the statistical analysis of the SA-ALT data was done.