软件确保智能测试用例生成PSO算法进展研究  

Overview of PSO for Automatic Test Case Generation in Software Assurance

作　　者：耿技[1] 聂鹏[1,2] 秦志光[1] 

机构地区：[1]电子科技大学计算机科学与工程学院,成都611731 [2]江西财经大学现代教育技术中心,南昌330013

出　　处：《电子科技大学学报》2012年第6期905-910,共6页Journal of University of Electronic Science and Technology of China

基　　金：国家自然科学基金(60973118;61133016);国家863计划(2011AA010706)

摘　　要：测试用例生成是软件测试的重要环节,是软件确保的重要组成部分,其中启发性算法是近年来测试用例自动生成研究领域的热点。对启发性测试用例生成的新方法PSO进行了介绍和分析,详细讨论了PSO算法适应度函数、PSO算法早熟与局部最优、种群规模对PSO算法的影响以及PSO参数优化问题,并将PSO与GA算法进行了对比分析。展望了PSO测试用例生成算法的未来研究方向,指出PSO测试用例生成算法目前应重点解决测试用例规模优化、早熟抑制和参数优选等问题。The automatic test case generation is a key phase of software testing and an important part ot software assurance. The study on the heuristic algorithms is an emerging area of the automatic test case generation in recent years. The new heuristic algorithm of PSO for the test case generation is reviewed and analyzed. Key issues on the PSO test case generation are discussed, including PSO fitness functions, PSO premature convergence and local optimum, swarm size impact, and parameter optimization. A contrastive analysis of PSO and GA in software testing is presented in detail. Finally, the future development of PSO test case generation algorithms is prospected including the test case swarm size optimization, the premature restraining, and the parameter optimization.

关 键 词：启发性算法 粒子群优化 软件确保 软件测试 测试用例生成 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]