考虑环境因素的软件可靠性增长模型

摘要

The failure intensity function will be misjudged in Software Reliability Growth Model (SRGM), because of the difference between the testing environment and the user environment. A logarithmic Poisson model considering environment factor was proposed, which was based on the M-O logarithmic Poisson execution time model, the representative of Musa execution-time models. It can represent the change law of the failure intensity function better, and the parameter estimation equations had been provided. The experimental results based on the failure data sets show that the proposed model has better fitting curve than some of other models.%软件可靠性增长模型中由于测试阶段和实际运行阶段环境的不同导致了失效强度函数的判断偏差.在Musa执行时间模型中的经典模型M-O对数泊松执行时间模型基础上,提出考虑环境因素的对数泊松模型,该模型能较好的刻画失效强度函数变化规律,并给出参数估计公式.通过对失效数据集的实验,结果表明该模型具有较好的拟合效果.