基于使用模型和遗传算法的测试数据自动产生技术

兵工学报 ›› 2006, Vol. 27 ›› Issue (6): 1051-1055.

基于使用模型和遗传算法的测试数据自动产生技术

周献中¹，孙勇成²，江金龙³

1．南京大学工程管理学院，江苏南京210093; 2．防空兵指挥学院，河南郑州450052 3．南京理工大学自动控制系，江苏南京210094

收稿日期:2005-12-20 上线日期:2014-12-25
通讯作者: 周献中

Automatic Test Data Generation Based on Usage Model and tjrenetic Algorithm

ZHOU Xian-zhong¹, SUN Yong-cheng²，JIANG Jin-long³

1. School of Management and Engineering, Nanjing University, Nanjing 210093, Jiangsu, China; 2. College of Air Defense Forces Command, Zhenzhou 450052，Henan, China; 3. Department of Automation, Nanjing University of Science and Technology, Nanjing 210094, Jiangsu, China

Received:2005-12-20 Online:2014-12-25
Contact: ZHOU Xian-zhong

摘要/Abstract

摘要： 测试数据的自动产生技术是软件测试的一个重要研究领域，是提高软件测试效率和效果的有效方法。为保证被测对象中实际运行更频繁的部分得到更充分的测试，把使用模型和遗传算法相结合，以整个被测试对象所需测试数据量为优化目标，给出了相应的规划模型，并提出了路径权重和目标路径集的确定方法，设计的适应性函数综合考虑了路径权重、路径覆盖率和路径接近程度三要素。实例证明了模型与方法的合理性及有效性。

关键词: 计算机软件 , 软件测试 , 测试数据 , 使用模型 , 遗传算法

Abstract: Automatic generation technology of test data is an important field of software test, which is an effective method to promote the efficiency and the effect of software test. In order to ensure that the more frequently used features would be tested more thoroughly, usage model and genetic algorithm were integrated together. One programming model was proposed, in which the optimal objective func?tion was defined as the required test data number for whole software test. The methods to decide the weight of path and to select the target path set were presented. Three key elements, 1.e. path weight, path coverage ratio and path approximation level were incorporated for designing the fitness function. An example shows the rationality and validity of the model and the methods.

Key words: computer software , software test , test data , usage model , genetic algorithm

中图分类号:

TP311.5

周献中，孙勇成，江金龙. 基于使用模型和遗传算法的测试数据自动产生技术[J]. 兵工学报, 2006, 27(6): 1051-1055.

ZHOU Xian-zhong, SUN Yong-cheng，JIANG Jin-long. Automatic Test Data Generation Based on Usage Model and tjrenetic Algorithm[J]. Acta Armamentarii, 2006, 27(6): 1051-1055.

参考文献

[1] Clarke J，Dolado J J, Harman M, et al. Reformulating software engineering as a search problem [ J ]. IEEE Proc on Software,2003,150(3):161-175.
[2] Wegener J, Baresel A, Sthamer H. Evolutionary test environ?ment for automatic structural testing[j], Information & Software Technology, 2001,43 (14) ： 841 - 854.
[3] Pe M，Goodman E D, Gao Z, et al. Automated software test da?ta generation using genetic algorithms [ J ]. IEEE Trans on Soft?ware Engineering, 1994，16 (8) :870 - 879.
[4] Jones B，Sthamer H, Eyres D. Automatic structural testing using genetic algorithms [ J ] . Software Engineering Journal, 1996，11(5):299-306.
[5] Sthamer H H. The automatic generation of software test data us?ing genetic algorithms [ D ]. Wales： University of Glamorgan, Pontyprid, 1996： 300-309.
[6] Tracey N, Clark J, Mander K. An automated framework for structural test data generation[C] . Proc of the 13th IEEE Con?ference on Automated Software Engineering, Hawaii，USA, 1998:285-288.
[7] Lin J C, Yeh P L. Using genetic algorithms for test case genera?tion in path testing [C].Proc of the Asian Test Symposium, 2000:241-246.
[8] Hermadi I, Ahmed M A. Genetic algorithm based on test data generator[C]. CEC’03，2003:85-91.
[9] Wegener J, Sthamer Ⅱ. Testing the temporal behavior of real?time tasks using extended evolutionary algorithms [ C ]. Eu- roSTAR, 1999： 270-271.
[10] 荚伟，谢军阊，奚红宇，等.遗传算法在软件测试数据生成中的应用[J].北京航空航天大学学报，1998，24(4): 434 - 437.
JIE Wei, XIL Jun-kai, XI Hong-yu，et a1. uenetic algorithms and its application in software test data generation[J]. Journal of Beijing University of Aeronautics & Astronautics, 1998，24(4)： 434 - 437. (in Chinese)
[11] Poore J H, Trammell C J. Application of statistical science to testing and evaluating software intensive systems[J]. Statistics, Testing & Defence Acquisition, 1998:40 — 57.
[12] Walton G H, Poore J H, Trammell C J. Statistical testing of software based on a usage modelLj」.Software Practice & Expe?rience, 1995, 25 (1):97-108.
[13] Trammell C.Quantifying the reliability of software： statistical testing based on a usage model[C]. Proc of the IEEE Int Soft?ware Engineering Standards Symposium, 1995： 208 - 218.
[14] Semmel G S，Linton D G. Determining optimal testing times for Markov chain usage models[C] . Proc of the 1998 IEEE South?east Conference, 1998： I — 4.
[15] Adams E N. Optimizing preventive service of software products [J] - IBM Journal for Research & Development, 1984, 28( I) :2 -14.