基于逐幸存路径处理的测试用例集约简技术

Test case reduction technology based on per-survivor processing algorithm

针对基于智能算法的测试用例集约简技术存在的容易陷入局部收敛、过早熟等问题,提出一种基于逐幸存路径处理的测试用例集约简算法,该算法运用逐幸存路径处理算法的顺序网格搜索思路,将测试用例集进行二进制编码,使其构成顺序网格搜索的状态空间,在状态转移阶段将代码覆盖率和测试用例有效执行时间作为分支度量,进而选择分支重量最大的路径作为幸存路径,从而剔除冗余状态,完成测试用例集的约简。实验结果表明,在相同的实验环境下,与其他算法相比,该算法在具有较高冗余率的同时也保证了较高的检错率,在一定程度上降低了软件测试的复杂度,从而提高了软件测试的效率。

Aiming at the problems of test case reduction(TCR) technology based on intelligent algorithms that it is easy to fall into local convergence and premature maturity, this paper proposed a test case set reduction algorithm based on per-survivor processing(PSP). The algorithm used the sequential grid search idea of PSP algorithm to binary code the test case set to form the state space of sequential grid search. In the state transition stage, the algorithm took the code coverage and the effective execution time of test cases as branch metrics. According to the calculation results, the path with the largest branch weight was the surviving path, so as to eliminate redundant states and complete the reduction of test case set. The experimental results show that under the same experimental environment, compared with other algorithms, this algorithm not only has higher redundancy rate(RR), but also ensures higher error detection rate(EDR). This algorithm reduces the complexity of software testing to a certain extent and improves the efficiency of software testing.