基于指数环境因子函数的软件可靠性增长模型

绝大部分软件可靠性增长模型采取的软件可靠性标准都是测试可靠性，实际上，软件的测试剖面很难真实地反映运行剖面，软件运行环境和测试环境不可能一致。考虑测试环境与运行环境差别，根据经验数据拟合环境因子，提出一种随时间变化的指数环境因子函数；并建立基于指数环境因子的非齐次泊松过程类软件可靠性增长模型；利用一组公开发表的数据集对模型进行评估。结果表明，本文提出的环境因子函数计算复杂度低，基于该函数的软件可靠性模型具有更好的拟合和预测能力。

Software reliability growth model with exponential environmental-factor function

The test and running environment can not be consistent. Considering the difference between test and running environment, a time-varying exponential environment factor function is proposed. The non-Homogeneous Poisson process software reliability growth model based on environment factor is established and evaluated using a group of public data. The results show that the proposed function has low computational complexity, the software reliability model based on this function has better fitting results and prediction ability.