Process and statistical analysis of interval censored data in HPM effect experiment
-
摘要: 在HPM效应实验中经常可以获得区间删失数据,为了能够合理利用这类数据对HPM效应进行有效分析,需要对它进行处理。根据电子器件的微波失效机理和实验现象,基于插值方法的思想,在充分利用删失数据信息情况下建立了不同阶插值精度的数据处理方法。理论分析可知,高阶精度处理方法要优于低阶精度方法。此外,根据构建的区间删失数据,通过统计分析可知,处理后数据与原始数据在统计意义上没有显著差异,可用于HPM效应研究,为数据的可靠分析提供了有利支撑。Abstract: The interval censored data(ICD) are often acquired by HPM effect experiment. These data must be processed to reasonably analyze microwave effect. Based on interpolation concept, two processing methods are established by using interval censored data information and the failure mechanism and experimental phenomenon of HPM effect. By theory analysis, the second order method is better than the first order method in interval. Based on given ICD, the processed data of those methods are not significantly different in statistics, and can be applied to studying the HPM effect.
-
Key words:
- HPM effect /
- interval censored data /
- data processing /
- statistical analysis
点击查看大图
计量
- 文章访问数: 1379
- HTML全文浏览量: 204
- PDF下载量: 452
- 被引次数: 0