【摘要】 现有测试用例集约简方法基本只考量优化集的规模和错误覆盖率,很少考虑其测试效率,导致优化集的测试运行时间较长,且运行代价较大。针对该问题,提出一种三目标搜索算法。对测试用例覆盖度和运行代价的双目标优先级模型进行细化,给出每个目标的定量细化方法,根据此模型计算出各个用例的优先级,运用到改进蚁群算法中作为各节点的初始信息素值进行优化遍历搜索,将错误检测率作为蚂蚁挑选下一个节点的影响因子,并制定信息素更新规则,使其能够尽快找到最优解。实验结果表明,与贪心算法、基本蚁群算法等相比,该算法得到的测试用例最小集规模较小,具有较高的错误检测率。更多还原

【Abstract】 Existing test case reduction methods basically only take the size and the error detection rate of the optimization set into consideration,rarely paying attention to the test efficiency,so that the test of the opitimization set takes long,and the test cost is high. For this problem,this paper proposes a triple target search algorithm. It details the double target priority model based on the coverage and the test cost of the test case,refines each target’s quantitative method,then calculates the priority of each test case according to the model,uses the priority as the initial pheromone value in the improved ant colony algorithm for ergodic optimization,at last regards the error detection rate as the impact factor for the ant to pick the next node,and formulates the pheromone update rule to find the optimal solution as soon as possible. Experimental results showthat compared with the greedy algorithm,basic ant colony algorithm and other algorithms,this algorithm can get a smaller optimization test suite,higher error detection rate.更多还原