摘要
旅行商问题(TSP)是组合优化问题的典型代表,针对TSP的求解提出一种离散型细菌觅食(DBFO)算法。该算法通过结合2-opt算法设计了一种适合处理离散型变量的趋化算子,将细菌觅食算法推广到了离散情形。同时,结合TSP的特点,在迁徙算子中引入基因库的思想来指导新个体的生成,提高了算法的搜索效率。通过对TSPLIB标准库中22个实例进行仿真实验。实验结果表明,该算法能够有效求解城市规模500以下的TSP,与混合蚁群算法和离散型萤火虫群算法相比,具有更好的全局收敛性和稳定性。
Traveling salesman problem（ TSP） is a typical representative of combinatorial optimization problems. This paper proposed a discrete bacteria foraging optimization（ DBFO） algorithm to tackle the TSP. It desinged a new chemotaxis operator which was combined with 2-opt algorithm so as to handle discrete variables,and it made the bacteria foraging algorithm extend to the discrete situation. With the characteristics of the TSP,it designed a new elimination operator which is combined with gene pool so as to direct the generation of new individuals,which improved the search efficiency. Simulation experiments on22 TSP problems from TSPLIB show that,the proposed algorithm can solve the TSP which city scale is below 500 effectively,and is superior to hybrid ACO and discrete GSO algorithm.
出处
《计算机应用研究》
CSCD
北大核心
2014年第12期3642-3645,3650,共5页
Application Research of Computers
基金
国家自然科学基金资助项目(71271034)
辽宁省科技重大项目(2011219009)
辽宁省教育厅科学研究一般项目(L2012173)
关键词
离散型细菌觅食优化算法
旅行商问题
2-opt
基因库
discrete bacteria foraging optimization algorithm
traveling salesman problem（TSP）
2-opt
gene pool
