改进的局部ART测试算法研究  

Research on Advanced Adaptive Random Testing by Localization

作　　者：孙林[1] 周兴社[1] 董云卫[1] 张凡[1] 陈俊文[1] 

机构地区：[1]西北工业大学计算机学院,西安710072

出　　处：《计算机科学》2008年第9期192-194,共3页Computer Science

基　　金：国家863项目(2006AA012162)

摘　　要：局部ART算法虽然克服了测试样本点与样本生成区域已测试顶点有可能距离很近的问题,但它对生成区域周围的已测试样本点对样本选择的影响考虑不足。本文针对局部ART算法的缺陷,提出了改进的局部ART算法,将生成区域周围一定范围内的已测试顶点都加入到临近已测试样本集中,在保持时间复杂度的同时,使测试样本的分布更加均匀。实验结果表明,改进的局部ART算法的错误检测能力得到了提升。Although Adaptive Random Testing has solved the problem that the generated test case may be too close to the nearby executed vertexes of the test generating zone, it does not consider the influences of the test cases near the test generating zone. This paper proposes the Advanced Adaptive Random Testing by Localization algorithm, which adds all the executed test cases near the zone to the nearby executed test set and thus spreads the test cases more evenly while maintaining the complexity. The experiment results show that the failure detecting ability of the Advanced ART by Localization is promoted.

关 键 词：改进的局部ART算法 局部ART算法 D-ART RRT 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构] TP333[自动化与计算机技术—计算机科学与技术]