摘要
研究遗传算法和蚁群算法可作为新兴的智能优化算法,在解决多目标、非线性的组合优化问题上表现出了传统优化算法无可比拟的优越性。基于将两种智能优化算法动态融合的思想提出了一种新的遗传蚁群算法(GA-ACO)。与已有的将遗传算子引入蚁群算法的结合方式不同之处在于,GA-ACO算法第一阶段采用了遗传算法生成初始信息素分布,在第二阶段采用蚁群算法求出最优解,从而有效地结合了遗传算法的快速收敛性和蚁群算法的信息正反馈机制。仿真结果表明,在具有深度陷阱的特殊障碍物环境下,应用GA-ACO算法求解机器人路径规划问题可以得到较好的的结果。
As two popular intelligence optimal algorithm, Ant Colony Optimal Algorithm （ACO） and Genetic Algorithm （GA） perform better than traditional optimal algorithms in many fields. Based on the integration of the two algorithms, the paper proposes a kind of GA - ACO Algorithm. Differ from the existent method of inserting genetic operators into ACO, the basic thought of the GAAC algorithm is that, the algorithm uses GA to generate the initial pheromone distribution in the former stage and then uses ACO to work out the final solution in the later stage. By doing so, the merits of the two algorithms can be made the most use of to obtain the efficiency of time and precision. The simulation result indicates that introducing GA - ACO algorithm to mobile robotic path planning problem in obstacle enviroment with deep traps illuminates its efficiency.
出处
《计算机仿真》
CSCD
北大核心
2010年第3期170-174,共5页
Computer Simulation
基金
国家自然科学基金(60674015)
上海市重点学科项目(B504)
关键词
遗传算法
蚁群算法
机器人路径规划
Genetic algorithm
Ant colony optimal algorithm
Mobile robotic path planning
