摘要
结合遗传算法(GA)的并行搜索结构和模拟退火(SA)的概率突跳性,并结合使用自适应的交叉算子和变异算子,提出了一种高效的自适应的SAGA混合优化算法。在自主开发的结构性测试工具WBoxTool中,使用自适应SAGA混合优化策略进行测试数据自动生成,并通过实例对基本遗传算法、自适应遗传算法和自适应SAGA进行了比较,结果表明自适应SAGA具有更强的搜索能力,可以更快的发现全局最优解。
Combining the parallel searching structure of genetic algorithm (GA) with the probahiliatic jumping property of simulated annealing (SA), also with the using of adaptive crossover operator and mutation operator, the adaptive SAGA hybrid optimization strategy is proposed. The adaptive SAGA hybrid optimization strategy was used to generate test data automatically in the structural test tool WBoxTool, which was developed on an independent basis. By comparing adaptive SAGA algorithm with the simple genetic algorithm and adaptive genetic algorithm on effectiveness, the adaptive SAGA algorithm has more strong searching ability that can abandon the local optimal solution and find the global one more quickly.
出处
《微电子学与计算机》
CSCD
北大核心
2006年第8期10-13,16,共5页
Microelectronics & Computer
基金
国家高技术研究发展计划(2003AA1Z2610)
关键词
测试数据自动生成
模拟退火算法
遗传算法
自适应
SAGA
Automatic test data generation, Simulated annealing algorithm, Genetic algorithm, Adaptive, SAGA
作者简介
郭斌,男,(1980-),硕士。研究方向为软件工程、智能优化。