构件集成软件测试用例生成智能算法分析  

Analysis on Intelligent Algorithms in Test Cases Generation for Component-Based Software

作　　者：张毅坤[1] 胡先智[1] 崔颖安[1] 夏辉[1] 张保卫[1] 

机构地区：[1]西安理工大学计算机科学与工程学院,陕西西安710048

出　　处：《微电子学与计算机》2008年第12期114-117,121,共5页Microelectronics & Computer

基　　金：陕西省自然科学基金项目(2005F07);陕西省教育厅专项基金项目(06JK230)

摘　　要：构件集成软件测试用例生成是测试中的重要的难题之一.提出了构件集成软件的结构树路径覆盖准则,研究智能算法的参数编码设计、适应度函数的构造、关键算子的设定.以三角形分类判断的构件集成软件为例仿真实验,重点从覆盖率、收敛情况、测试用例数目等方面对各智能算法进行比较分析.结果表明,融合的遗传算法与蚂蚁算法与其他算法相比有一定优势.Test case generation for component integration software is one of the important problems of software testing. This paper is based on the layer-tree path coverage criteria for component integration software. The authors research on coding design of parameters, the construction of fitness function and pivotal operator definition of four different intelligent algorithms. Simulating experiments on the basis of component integration software of the Triangle classification, through comparison and analysis of different algorithms with the aspects of coverage ratio, convergence condition and test cases quantities, the experimental results show that the combination of Genetic and Ant Algorithm is better than the others.

关 键 词：智能算法 结构树路径覆盖 构造适应度函数 关键算子 测试用例生成 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]