摘要
为了提高遗传算法对满意解的搜索和优化能力,采用基于模糊逻辑的自适应控制策略,提出了一种符号编码的自适应遗传算法。该算法可自动均衡搜索和优化关系,采用6个模糊控制器实现对选择、交叉、变异操作的动态参数组合控制。试验和理论分析表明,六模糊控制器的组合控制方式可以综合两模糊控制器或三模糊控制器独立控制的性能。对旅行商(TSP:TravelingSalesmanProblem)问题的求解结果表明:该算法在解决类似于TSP的组合优化问题时,具有比标准遗传算法更好的性能。
Adaptive control strategies based on fuzzy logic are used to enhance the GA(Genetic Algorithm)'s capability of exploration and optimization for satisfying solution. Fuzzy logic controllers have been built for inducing suitable exploration and optimization relationship throughout the run of GA automatically. The proposed symboliccoded AGA(Adaptive Genetic Algorithm) uses 6 fuzzy logic controllers (6FLCs) to control genetic operating parameters of selection, crossover and mutation dynamically. Experiments and theoretic analyses show that the combination control method of 6FLCs could combine the good behaviors of the 2FLCsbased AGA and the 3FLCsbased AGA as a whole. Experimental results of the TSP(Traveling Salesman Problem) demonstrate that the 6FLCsbased AGA is more efficient than a standard GA in solving combinatorial optimization problems which is similar to the TSP.
出处
《吉林大学学报（信息科学版）》
CAS
2003年第4期329-333,共5页
Journal of Jilin University（Information Science Edition）
基金
国防基础科研基金资助项目
关键词
自适应遗传算法
模糊控制器
动态参数控制
旅行商问题
Adaptive genetic algorithm
Fuzzy logic controller
Dynamic parameters control
Traveling salesman problem
