基于自适应粒子群优化的组合测试用例生成方法

摘要

最小覆盖表生成是组合测试研究的关键问题.基于演化搜索的粒子群算法在生成覆盖表时能得到较优的结果,但其性能受配置参数的影响.针对此问题,将one-test-at-a-time策略和自适应粒子群算法相结合,以种群粒子优劣为依据对惯性权重进行自适应调整,使其在覆盖表生成上具有更强的适用能力.为进一步提升算法性能,构造了一个优先级度量函数用于度量每个组合的权值,优先选取权值最高的组合用于单条测试用例的生成.最后,编程实现该算法,并将其与原有粒子群算法在组合测试用例集生成上展开对比性实验分析,结果证实该算法在规模和执行时间上具有竞争力.%Obtaining minimum coverage array is one of the key issues in the combination test.Particle swarm optimization(PSO),as one of the evolutionary search based methods,can obtain the smallest covering arrays,but its performance is significantly impacted by the parameters.To solve this problem,we combined one-test-at-a-time strategy and particle swarm optimization and proposed an adaptive particle swarm optimization algorithm.Based on the quality of the particles in the population,the strategy adaptively adjusts inertia weights which makes it have stronger ability of application.In order to further improve the performance of the algorithm,we constructed a priority measure function which is used to measure the weight of each combination,and we preferred to select a combination which has the highest weight to generate a single test case.Finally the paper implemented the algorithm by programming,and compared this approach with the original particle swarm optimization algorithm in test suite size and generation time.The results show the competitiveness of this approach.