软件可靠性增长G-O模型的改进

Improvement of software reliability growth G-O model

在传统的软件可靠性增长G-O模型中,故障检测率和初始的故障总数是影响软件可靠性的2个重要因素。为了提高软件可靠性评估的可信性,考虑到在软件纠错的过程中可能会引入新的错误,把模型中潜在的故障总数和故障检测率看作随时间变化的函数,提出了改进的G-O模型,给出了解析方法,并将改进前后的G-O模型进行了对比,通过实例进行了验证。

Software fault detection rate and total number of initial faults are two important factors that influence software reliability in the traditional software reliability growth G-O model.In order to increase the reliability,the fact that new errors may be introduced during error correction process of software is considered.The total number of potential faults and fault detection rate are defined as a function of time,an improved G-O model is proposed,and its analytical method is given.The G-O model before and after improvement are compared and verified by some examples.