基于统计分析的自适应蚁群算法及应用  

SELF-ADAPTIVE ANT COLONY ALGORITHM BASED ON STATISTICAL ANALYSIS AND ITS APPLICATION

作　　者：许明乐 游晓明[1] 刘升[2] 

机构地区：[1]上海工程技术大学电子电气工程学院,上海201620 [2]上海工程技术大学管理学院,上海201620

出　　处：《计算机应用与软件》2017年第7期204-211,共8页Computer Applications and Software

基　　金：国家自然科学基金项目(61075115;61403249);上海市教委科研创新重点项目(12ZZ185)

摘　　要：路径规划是机器人关键技术之一。利用改进的蚁群算法进行机器人的路径规划。针对传统蚁群算法收敛速度慢且易陷入局部最优解的缺陷,在Ant Colony System算法基础上,对每代蚁群动态随机统计分析,提取最优、平均和最差的蚂蚁信息,构成自适应算子用于局部信息素的自适应更新。仿真实验结果证明该自适应算子在平衡增加收敛速度和陷入局部最优解矛盾的问题中是有效的。Path planning is one of the key technologies of robot. In this paper, the improved ant colony algorithm is applied to robot path planning. Aiming at the shortcoming of traditional ant colony algorithm which is slow to converge and easy to fall into local optimum, the dynamic random statistical analysis of each ant colony is performed based on the Ant Colony System algorithm. The optimal, average and worst ant information are extracted to form an adaptive operator for the local pheromone adaptive updating. Simulation results show that the proposed adaptive operator is effective in solving the problem of increasing the convergence speed and falling into the local optimal solution.

关 键 词：路径规划 自适应精英策略 统计分析 蚁群算法 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]