基于遗传算法的测试用例生成工具研究

Tool for Automated Test Data Generation Based on GA

以Korel的遗传算法生成测试用例策略为基础,在适应函数、染色体编码、遗传算法效率等方面提出了相应的算法改进。引入程序结构依赖系数来确定个体的适应值,使得整个搜索更加智能化;解决编码长度预测问题;采用改进的自适应遗传算法提高用例的搜索效率。同时对实现测试用例生成过程中的关键技术进行了分析。最后结合算例对软件进行了验证。

This paper proposes an algorithms improved on fitness function, chromosome size and the efficiency of GA on the basis of Korel's method. The dependency coefficient of program structure is introduced. The dynamic variable coding length of chromosome is adopted, and the efficiency searching test cases is enhanced. At same time, the key technology for implementing automated test cases is also discussed. Finally the tool software is verified by an example.