摘要
针对传统蚁群算法存在搜索效率低、耗时长、易发生早熟收敛等不足,采用"最大-最小蚂蚁"思想,抑制信息素增长过快而导致的停滞现象;采用新的概率选择模式,避免死锁现象的发生;采用转轮赌法选择蚂蚁的行走栅格,保证栅格的选择概率;采用"精英蚂蚁"策略,加强循环中最优路径对蚂蚁的吸引力,并用优化排序策略优化精英蚂蚁;当所有蚂蚁完成一次搜索后,引入"最优-最差蚂蚁"思想,削弱当前最差路径,增加其信息素的更新。仿真结果表明,改进后的算法更加省时、高效,有效提高了机器人的路径规划性能。
Aiming at the deficiencies of traditional ant colony algorithm, such as low searching efficiency, long time-consuming and prone to premature convergence, this algorithm adopts the idea of ＂maximum- minimum ant＂ to restrain the stagnation phenomenon caused by the rapid growth of pheromone. The new probability selection mode is adopted to avoid the occurrence of deadlock phenomenon, and the selection of Ant＇s walking grid is ensured by the method of wheel gambling, the selection probability of the grid is guaranteed, the optimal path of the circle is strengthened by using the ＂elite ant＂ strategy, and the elite ants are optimized by optimizing the sorting method. When all ants complete a search, the idea of ＂best- worst ant＂ is put in place to weaken the current worst path and increase its pheromone update. The simula- tion results show that the improved algorithm is more the path planning performance of the robot. time-saving and efficient, and effectively improves
作者
周敬东
郑小玄
ZHOU Jingdong;ZHENG Xiaoxuan(Research and Design Institute of Agricultural Machinery,Hubei Univ.of Tech.,Wuhan 430068,China)
出处
《湖北工业大学学报》
2018年第5期49-52,101,共5页
Journal of Hubei University of Technology
基金
"十三五"国家重点研发计划智能农机装备专项(2017YFD0700905)
关键词
蚁群算法
路径规划
步行机器人
ant colony algorithm
path planning
walking robot
