摘要
利用多组软件失效数据集分析了G-O模型参数的稳定性,指出该模型参数的稳定性较差.提出了再校准G-O模型的参数以提高可靠性预测的准确性,建立了再校准的G-O模型-RGOM,对这两个模型的性能进行了对比分析.结果表明,RGOM能获得较好的预测效果.
Some software failures data sets have been examined to analyze stability of the parameters of a G -O model. The poor stability of the parameters prevents the model from giving an accurate prediction of software failures during a given mission period. Recalibrating the parameters is proposed and a recalibrated G -O model called RGOM is put forward. A software failures data set was used to compare the performance of the two models. It shows that RGOM can give a more accurate result in that data set.
出处
《哈尔滨工业大学学报》
EI
CAS
CSCD
北大核心
2005年第12期1612-1614,共3页
Journal of Harbin Institute of Technology
基金
总装备部"十五"预研基金资助项目(41316.1.2)
