基于优化蚁群算法的软件测试数据自动生成方法

Automatic Generation Method of Software Test Data Based on Optimized Ant Colony Algorithm

为提高测试数据自动生成的可行性和检索性能,提出基于优化蚁群算法的软件测试数据自动生成方法。构建软件测试数据自动生成模型,通过程序分析单元构建测试路径集合;算法解算单元采用蚁群算法,编码处理输入变量生成初始蚁群,通过蚂蚁搜索路径、释放信息素、编码更新等过程获取测试数据;驱动程序单元获取测试数据后,调用桩程序分析数据有效性,并反馈结果。最后利用遗传算法优化蚁群算法,对获取的路径实施交叉操作,重组搜索子域,提升所生成软件测试数据的多样性。实验结果显示该方法可改善局部最优解问题,具有较好的搜索性能。

In order to improve the feasibility and retrieval performance of automatic test data generation,a software test data generation method based on optimized ant colony algorithm is proposed.It builds the automatic generation model of software test data,and builds the test path set through the program analysis unit.The algorithm solution unit uses the ant colony algorithm,codes and processes the input variables to generate the initial ant colony,and obtains the test data through the process of ant searching the path,releasing pheromones,and updating the code.After the driver unit obtains the test data,it calls the pile program to analyze the validity of the data and feedback the results.Finally,the genetic algorithm is used to optimize the ant colony algorithm,implement cross operation on the obtained path,reorganize the search sub domain,and improve the diversity of the generated software test data.The experimental results show that this method can improve the local optimal solution problem and has good search performance.