摘要
为了提高机器人路径规划的速度,提出一种全新的机器人路径规划算法.算法中,青蛙以随机方式和启发方式两种策略从可选栅格集中选择栅格.子蛙群进行更新时,最坏青蛙根据与子群最优青蛙或全局最优青蛙的路径交点栅格更新路径.为了进一步提高搜索速度,算法中引入评分法,只对得分小于阈值的青蛙进行更新,同时采用双种群双向搜索的方法.大量仿真实验结果表明,该算法比同类算法的收敛速度提高数十倍以上,能在复杂的静态障碍环境中,迅速规划出一条安全避碰的优化路径.
In order to obtain a higher convergence speed,a new robot path planning algorithm is proposed. In this algorithm,frogs choose grids from the optional grid sets either randomly or heuristically. The worst frog updates its path according to the grids which intersect the paths of the best sub-group frogs or the global optimal frog. To further speed up the searching,a scoring method and a search mechanism of bi-directional frog-swarms is introduced. Then the frog＇s path will be updated only when corresponding score is under a threshold. A large number of simulation experiments show that the algorithm convergence speed is several times faster than similar algorithms＇. Moreover,it can avoid collision and plan an optimal path rapidly in a complex static obstacle environment.
出处
《小型微型计算机系统》
CSCD
北大核心
2014年第7期1631-1635,共5页
Journal of Chinese Computer Systems
基金
国家自然科学基金项目(60673102
61073118/F020508)资助
江苏省高校自然科学基金项目(10KJD520004)资助
关键词
机器人
路径规划
蛙跳算法
双向搜索
mobile robot
path planning
shuffled frog leaping algorithm
Bi-directional search
