基于并行化粒子群算法的组合测试用例生成  

A parallel test case generation method for pairwise combination

作　　者：王曙燕 张海清 孙家泽 WANG Shuyan;ZHANG Haiqing;SUN Jiaze(School of Computer Science and Technology,Xi'an University of Posts and Telecommunications,Xi'an 710121,China)

机构地区：[1]西安邮电大学计算机学院

出　　处：《西安邮电大学学报》2019年第5期64-68,共5页Journal of Xi’an University of Posts and Telecommunications

基　　金：陕西省工业科技攻关计划资助项目(2018GY-014,2017GY-092);西安市科技计划资助项目(GXYD17.10)

摘　　要：针对粒子群算法生成组合测试用例消耗时间过长的问题,提出一种并行化粒子群算法生成两两组合测试用例的方法。基于大数据平台Spark,将全部需要被覆盖的两两组合进行分组,并下发到集群中各个节点上进行寻优操作;采用one-test-at-a-time策略与自适应粒子群算法相结合的方式进行寻优;待各个节点寻优结束后,利用Spark进行结果收集,并对收集后的用例集进行约简操作。实验结果表明,该方法有效地减少了生成两两组合测试用例集的消耗时间。To solve the problem that the combined test cases consume too long by the particle swarm optimization algorithm,a parallel particle swarm optimization algorithm is proposed to generate the two-two combined test cases.The method is based on the big data platform Spark.Firstly,all the two pairs of combinations that need to be covered are grouped and sent to each node in the cluster for optimization.In the optimization stage,one-test-at-a-time strategy is mainly used.The method is also combined with the adaptive particle swarm optimization algorithm.After each node is optimized,the Spark is used to collect the results,and thus the collected use case set is reduced.Experimental results show that this method can effectively reduce the consumption time of the pairwise combinatorial test case set.

关 键 词：两两组合测试 并行化粒子群算法 SPARK 覆盖表 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论]