基于优化Markov链使用模型的可靠性测试技术  

Software Reliability Testing Method Based on Optimizated Markov Chain Usage Model

作　　者：吴亮[1] 方建勇[2] 

机构地区：[1]海军驻大连地区军代表室,辽宁大连116021 [2]江苏自动化研究所,江苏连云港222061

出　　处：《指挥控制与仿真》2015年第5期124-127,138,共5页Command Control & Simulation

基　　金：总装备部十二五预研基金项目

摘　　要：Markov链使用模型是一种应用较为广泛的软件可靠性测试模型,但是如何生成使用模型中的转移概率仍然是一个亟待解决的问题。传统Markov链模型中的转移概率往往依赖用户、开发者及领域专家给出一个具体的数值,是非常不合理的。因此,提出了一种使用三参数区间数表示转移概率的方法,对Markov链模型进行优化并给出可靠性测试用例的生成方法。实验表明,使用三参数区间数构造出的软件使用模型要优于传统Markov链,并可以生成较为符合实际情况的可靠性测试用例,从而有助于提高可靠性评估的准确性。Markov chain usage model is a widely used model in software reliability testing,and how to assignment the migrating probability between states of Markov chain usage model is still a problem to be solved. In traditional Markov chain model,the probability is often given as a fixed value by software users,developers and domain experts,and this approach is unreasonable. In order to deal with this problem,a migrating probability assignment mothod based on three-parameter interval number is proposed to build the Markov chain usage model and the method for generating test cases is also given. As the testing result shows,the software usage model based on the three-parameter interval number is better than the traditional Markov chain model and it is helpful to improve the accuracy of the software reliability assess.

关 键 词：MARKOV链 软件可靠性测试 三参数区间数 测试模型 

分 类 号：TP311.55[自动化与计算机技术—计算机软件与理论]