基于改进遗传算法的接口测试数据集的生成方法
|
摘要
针对用电信息采集统一接口平台需要大量包含足够测试用例占比的测试数据这一难题,提出基于改进遗传算法的用电信息采集系统统一接口平台测试数据集的生成方法;在对原始数据预处理的基础上,采用基于相似度的交叉算子对原始群体扩充,使用提出基于群体趋势不变的染色体变异算法,在保持群体数据集特性的情况下增大测试用例占比,从而形成测试数据的自动生成方法;应用该测试数据集生成方法,基于某省级电力公司2016年7—9月300万个典型用电客户用电数据进行数据集生成实验,利用熵原理比较无变异因子、插值法变异和改进遗传算法分别生成的测试数据与原始数据的重合度。结果表明,改进遗传算法生成的测试数据集,具有同用电信息采集系统采集数据相同的属性和属性值分布以及类似的属性关联关系,能够满足测试用例需求。
In order to solve the problem of test data generation in the unified interface platform of power energy data acquire,the method of generating the test data set based on the improved genetic algorithm was proposed. On the basis of the original data preprocessing,the crossover operator based on similarity was used to expand the original group. The algorithm of chromosome mutation based on group trend invariance was proposed to increase the test case concentration under the condition of maintaining the characteristics of the group data set,thus forming the automatic generation method of test data. A data set generation experiment was carried out based on the typical power data from July to September 2016 of one typical provincial power company by the test data set generation method. The entropy principle was used to compare the coincidence degree between the test data and the original data,using the principle of entropy to compare the non variant factor,the interpolation method variation and the use of this method. The experimental results show that the test data set generated by this method has the same attribute and attribute value distribution,similar attribute relation with the power energy data acquire system,and meets the requirement of test case.
In order to solve the problem of test data generation in the unified interface platform of power energy data acquire,the method of generating the test data set based on the improved genetic algorithm was proposed. On the basis of the original data preprocessing,the crossover operator based on similarity was used to expand the original group. The algorithm of chromosome mutation based on group trend invariance was proposed to increase the test case concentration under the condition of maintaining the characteristics of the group data set,thus forming the automatic generation method of test data. A data set generation experiment was carried out based on the typical power data from July to September 2016 of one typical provincial power company by the test data set generation method. The entropy principle was used to compare the coincidence degree between the test data and the original data,using the principle of entropy to compare the non variant factor,the interpolation method variation and the use of this method. The experimental results show that the test data set generated by this method has the same attribute and attribute value distribution,similar attribute relation with the power energy data acquire system,and meets the requirement of test case.
引文
[1]李程,魏强,彭建山,等.基于分解重构的网络软件测试数据生成方法[J].计算机科学,2013,40(10):108-113.
[2]赵诗奎,王林瑞,石飞,等.作业车间调度问题综述[J].济南大学学报(自然科学版),2016,30(1):74-80.
[3]马志兵,余荣学,周丽平.基于遗传算法的测试用例自动生成技术研究[J].电子技术与软件工程,2015(6):193-194.
[4]许力,江勇.基于遗传算法的数据流测试用例自适应生成算法[J].计算机系统应用,2013,22(7):90-94.
[5]姜元鹏,董庆杰.基于遗传算法和分支覆盖的测试数据生成方法[J].计算机工程与设计,2016,37(1):112-117.
[6]冯俊池,于磊.测试数据生成中遗传算法的改进[J].计算机辅助设计与图形学学报,2015,27(10):2008-2014.
[7]周小飞,赵瑞莲,李征.基于多种群遗传算法的可扩展有限状态机测试数据自动生成[J].计算机应用与软件,2015,32(11):1-6,52.
[8]王玲,秦克云,刘雅雅.区间值模糊软集的相似度及距离和熵[J].济南大学学报(自然科学版),2015,29(5):361-366.
[9]王建民.基于维持种群多样性的测试数据生成算法的研究[J].计算机研究与发展,2012,49(5):1039-1048.
[10] VARSHNEY S,MEHROTRA M. Search-based test data generator for data-flow dependencies using dominance concepts,branch distance and elitism[J]. Arabian Journal for Science&Engineering,2016,41(3):1-29.
[11]霍远秦.基于遗传算法的多路径测试数据自动生成[D].北京:北京化工大学,2013.
[12] YAO X J,GONG D W,ZHANG G J. Constrained multi-objective test data generation based on set evolution[J]. IET Software,2015,9(4):103-108.
[13]李圣文,凌微,龚君芳,等.一种基于熵的文本相似性计算方法[J].计算机应用研究,2016,33(3):665-668.
[2]赵诗奎,王林瑞,石飞,等.作业车间调度问题综述[J].济南大学学报(自然科学版),2016,30(1):74-80.
[3]马志兵,余荣学,周丽平.基于遗传算法的测试用例自动生成技术研究[J].电子技术与软件工程,2015(6):193-194.
[4]许力,江勇.基于遗传算法的数据流测试用例自适应生成算法[J].计算机系统应用,2013,22(7):90-94.
[5]姜元鹏,董庆杰.基于遗传算法和分支覆盖的测试数据生成方法[J].计算机工程与设计,2016,37(1):112-117.
[6]冯俊池,于磊.测试数据生成中遗传算法的改进[J].计算机辅助设计与图形学学报,2015,27(10):2008-2014.
[7]周小飞,赵瑞莲,李征.基于多种群遗传算法的可扩展有限状态机测试数据自动生成[J].计算机应用与软件,2015,32(11):1-6,52.
[8]王玲,秦克云,刘雅雅.区间值模糊软集的相似度及距离和熵[J].济南大学学报(自然科学版),2015,29(5):361-366.
[9]王建民.基于维持种群多样性的测试数据生成算法的研究[J].计算机研究与发展,2012,49(5):1039-1048.
[10] VARSHNEY S,MEHROTRA M. Search-based test data generator for data-flow dependencies using dominance concepts,branch distance and elitism[J]. Arabian Journal for Science&Engineering,2016,41(3):1-29.
[11]霍远秦.基于遗传算法的多路径测试数据自动生成[D].北京:北京化工大学,2013.
[12] YAO X J,GONG D W,ZHANG G J. Constrained multi-objective test data generation based on set evolution[J]. IET Software,2015,9(4):103-108.
[13]李圣文,凌微,龚君芳,等.一种基于熵的文本相似性计算方法[J].计算机应用研究,2016,33(3):665-668.