免疫粒子群算法的测试数据生成
Immune particle swarm optimization algorithm and its application on test data generation
焦重阳 1周清雷 2张文宁3
作者信息
- 1. 中国人民解放军战略支援部队信息工程大学,河南郑州 450001;数学工程与先进计算国家重点实验室,河南郑州 450001;郑州工业安全职业学院河南信息工程学校,河南郑州 450011
- 2. 郑州大学计算机与人工智能学院,河南郑州 450001
- 3. 中原工学院软件学院,河南郑州 450007
- 折叠
摘要
为有效改善粒子群算法进化后期收敛速度慢,克服易陷入局部极值的缺陷,提出一种自适应免疫粒子群算法并在面向路径的测试数据生成中得到应用.本文提出 自适应的惯性权重的调整方法和学习因子的调节策略,加快算法的搜索速率;引入免疫算法中的免疫算子,提出抗体的浓度调节机制,使得粒子群的多样性更加丰富,提升算法的寻优能力;通过免疫选择操作,避免算法的早熟收敛;以分支函数叠加法构造适应度函数.实验结果表明,该算法避免了粒子群算法早熟收敛现象的发生,有效地提高了测试数据自动生成的效率.
Abstract
To overcome the phenomena of falling into local optimal value and premature convergence of the standard particle swarm algorithm in the late stage of evolution,an optimization method of adaptive immune particle swarm optimization algorithm was proposed to generate path-oriented test data automatically.An adaptive adjustment scheme based on inertial weight and an adjustment scheme based on learning factor were proposed to improve the convergence speed of the algorithm.The regulation mechanism of antibody concentration was put forward to improve the diversity of the population and increase the search ability.Through the immune selection,the premature convergence of the algorithm was effectively avoided.The fitness function was constructed by the summation of branch functions to better evaluate the quality of the generated test data.Experimental results show that the proposed method can avoid premature convergence of particle swarm optimization algorithm and effectively improve the efficiency of generating test data automatically.
关键词
粒子群算法/测试数据生成/惯性权重/学习因子/免疫算子/种群多样性/免疫选择Key words
particle swarm optimization algorithm/test data generation/inertia weight/learning factor/immune operator/popu-lation diversity/immune selection引用本文复制引用
基金项目
国家自然科学基金面上项目(61572444)
出版年
2024
NETL
NSTL
维普
万方数据