基于适应度函数的联锁软件测试用例生成仿真  

Simulation of Interlocking Software Test Case Generation Based on Fitness Function

作　　者：刘明 王新强 LIU Ming;WANG Xin-qiang(School of Software and Communication,Tianjin Sino-German University of Applied Sciences,Tianjin 300350,China)

机构地区：[1]天津中德应用技术大学软件与通信学院,天津300350

出　　处：《计算机仿真》2022年第12期408-412,共5页Computer Simulation

摘　　要：为提高联锁软件的可靠性,避免逻辑运算出现失误,提出基于日志挖掘的联锁软件测试用例生成算法。将联锁系统软件分为人机对话、联锁运算与任务执行三个模块,分析不同模块执行的具体任务;采用支持向量机分类算法,清洗从联锁系统中采集的日志数据;通过频繁模式挖掘算法,根据支持度阈值挖掘用户日志,了解用户操作行为;分析挖掘结果生成测试项目,以语句覆盖和分支覆盖为生成准则;使用遗传算法,设计编码方式,设置适应度函数与遗传算子,构建测试用例生成模型。仿真结果表明,所提算法生成用例的漏检率与误检率较低,可节省测试开销,优化软件性能。In order to improve the reliability of interlocking software and avoid errors in logic operation, based on log mining, this paper puts forward an algorithm for generating interlocking software test cases. Firstly, we divided the interlocking software into three modules: man-machine conversation, interlocking operation and task execution, and then analyzed specific tasks executed by different modules. Secondly, we used a support vector machine classification algorithm to clean the log data collected from the interlocking system. According to the support threshold, we used the frequent pattern mining algorithm to mine user logs and thus to understand user operation behavior. Taking statement coverage and branch coverage as the generation criteria, we analyzed the mining results and created some test items. Moreover, we used the genetic algorithm, including designing the coding mode and setting fitness functions and genetic operators. Finally, we built a model for generating test cases. Simulation results show that the missing rate and false detection rate of the use cases created by the proposed algorithm are low, which can save test overhead and optimize software.

关 键 词：日志挖掘 联锁软件 测试用例 频繁模式挖掘 遗传算法 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]