软件测试数据自动生成方法的研究进展  

Research on progress of automated software test data generation method

作　　者：邓璐娟[1] 卢华琦[1] 张科德[1] 

机构地区：[1]郑州轻工业学院计算机与通信工程学院,河南郑州450002

出　　处：《郑州轻工业学院学报（自然科学版）》2009年第4期64-67,共4页Journal of Zhengzhou University of Light Industry:Natural Science

基　　金：河南省新世纪优秀人才支持项目(2005HANCET-03)

摘　　要：软件测试数据自动生成分为功能测试数据生成和结构测试数据生成.就目前的研究而言,功能测试数据生成方法包括边界值法、等价类划分法、决策表法、因果图法、功能图法、错误推测法、正交实验方法.结构测试数据生成方法包括随机测试数据生成方法、面向目标的测试数据生成方法、面向路径的测试数据生成方法、基于GA的结构测试数据生成方法.功能测试数据生成是按照需求规格说明设计测试数据的方法,测试简单,完全不考虑程序内部结构和内部特性.结构测试数据生成只根据程序的内部结构设计测试数据.在规划测试方案时,需要将两者结合起来应用,加入其他的智能算法是帮助测试数据生成的一个发展方向.Automatic generation of software test data can generally be divided into functional test data generation and the structure of the test data generation. To now study functional test data generation methods, including boundry value analysis method, equivalence class partition method, the decision-making table method, cause and effect diagram method, functional diagram method, error guessing method, the orthogonal experimental method. Structural test data generation methods, including random test data generation methods, goal-oriented test data generation methods, path-oriented test data generation methods, GAbased structural test data generation methods. Functional test data generation is a method which designed test data in accordance with the requirements specification and test is simple. Do not consider internal characteristics and intemal structure of the Procedure. Structural test data generation is a method which designed test data in accordance with internal structure of the prodrue. We need to combine them to application when planning the test program. And adding other intelligent algorithms will also help test data generated

关 键 词：测试生成 数据生成 遗传算法 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]