基于佳点集遗传算法的多路径覆盖测试用例生成  

Multi-path Coverage Test Case Generation Based on Improved Good Point Set Genetic Algorithm

作　　者：程孟飞 丁蕊[1] CHENG Mengfei;DING Rui(Mudanjiang Normal University,Mudanjiang 157011)

机构地区：[1]牡丹江师范学院,牡丹江157011

出　　处：《计算机与数字工程》2022年第9期1940-1944,共5页Computer & Digital Engineering

基　　金：国家自然科学基金项目“基于行为博弈论的连续组合拍卖策略及优化方法研究”(编号:61472095);黑龙江省教育厅重点项目“面向关系的多目标优化算法及其应用研究”(编号:1355ZD005);黑龙江省教育厅项目“基于群体智能优化算法的测试数据生成方法研究”(编号:1451TD003)资助。

摘　　要：面向路径覆盖的测试用例生成作为一项有效的白盒测试技术,对提高软件测试效率和降低测试成本有重要意义。针对传统遗传算法过早收敛和收敛速度慢的问题,提出一种基于佳点集遗传算法的多路径覆盖测试用例生成方法。算法根据问题特征采用不同的编码方式,对于二进制编码的个体,设计佳点集交叉算子生成子代个体,实数编码的个体,设计一种混沌交叉方法生成子代个体;考虑分支距离和个体穿越路径与目标路径矩阵的相似度作为适应度函数。实验结果表明,与其他算法相比,该方法在覆盖率以及测试时间方面均具有优势。As an effective white-box testing technology,path coverage-oriented test case generation is of great significance to improve software testing efficiency and reduce testing costs. Aiming at the problems of premature convergence and slow convergence of traditional genetic algorithm,a method for generating multi-path coverage test cases based on good point set genetic algorithm is proposed. The algorithm adopts different coding methods according to the characteristics of the problem. For individuals with binary coding,a good point set crossover operator is designed to generate offspring individuals. For individuals with real number coding,a chaotic crossover method is designed to generate offspring individuals. Considering branch distance and individual crossing,the similarity between the path and the target path matrix is used as a fitness function. The experimental results show that the method has advantages in both coverage and test time compared with other algorithms.

关 键 词：白盒测试 多路径覆盖 遗传算法 佳点集 混沌交叉 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]