【摘要】 非齐次泊松过程(Non—homogeneos Poisson Process)类模型简称(NHPP 在软件可靠性工程的软件可靠性预计工作中,有着广泛的应用.而且,通过实践的证明,用它们来描述非均匀性的软件测试过程,也是成功的.在应用 Ohba 的三参数 NHPP 模型时,由于参数声(错误引入率。β>0)的引入是从 G-O 模型演化来的,致使在以该模型拟合软件的故障数据时,产生了奇异性.为了解决在应用 Ohba 三参数 NHPP 模型中的这一奇异性问题,我们应用常用的路径跟踪方法,较好地克服了奇异性在参数估计过程中的不良影响.更多还原

【Abstract】 In software reliability .engineering, there is extensive application of Non-homogeneous Poisson Process models to software reliability predictions. And the factis proved by practice,it is sue- cessful to describe the non-homogeneous software test process of a software project by using these mod- els. In application process of Ohba’s three parameter non-homogeneous Poisson Process model, when we fit the software failure data set by using it, there is singularity originating from introduee of param- eter β, error introduce rate, β>0. We apply common path folowing method to overcome the singulari- ty successfully.更多还原