自适应人工鱼群算法在清罐移动机器人路径规划中的应用  

An Adaptive Artificial Fish School Algorithm for Path Planning of Mobile Tank-clearing Robot

作　　者：周利坤[1,2] 刘宏昭[1] 

机构地区：[1]西安理工大学机械与精密仪器工程学院,西安710048 [2]武警工程大学装备运输系,西安710086

出　　处：《机械科学与技术》2012年第7期1085-1089,共5页Mechanical Science and Technology for Aerospace Engineering

基　　金：国家火炬计划基金项目(07C26213711606);陕西省重点学科建设专项资金项目(102-00X903)资助

摘　　要：为了解决移动机器人路径规划智能方法易陷入局部最优问题,提出一种自适应变步长、变拥挤度因子和变视野域的自适应人工鱼群算法,并将该算法引入清罐移动机器人全局路径规划问题之中。针对人工鱼群算法的不足,通过自适应调整因子调整人工鱼的可视域、移动步长和拥挤度因子,使算法在清罐移动机器人路径规划中的遍历性得到改善,既可获得全局最优路径,又可以实现局部搜索,避免了传统人工鱼群算法局部寻优能力弱的缺点。仿真结果表明:基于自适应人工鱼群算法的清罐移动机器路径规划方法,能快速获得全局最优路径、提高算法的收敛速度和精度,与人工鱼群算法相比,具有收敛速度快,计算效率高的优点。To solve the problem that the intelligent mobile robot path planning method is easy to fall into local opti- mization, this paper presents an adaptive artificial fish school algorithm with variable step size, variable congestion factor and the variable view domain, for the global path planning of mobile cleaning tank robot. Aiming at the shortcomings of artificial fish school algorithm, the new algorithm uses an adaptive adjustment factor to adjust the artificial fish＇s visual field, moving step and the congestion factor to improve the periodicity in cleaning tank path planning of mobile robot, and can get the global best optimal path and local search to avoid the traditional local searching AFSA weak shortcomings. Simulation results show that compared with the artificial fish school algorithm, the adaptive artificial fish school algorithm can quickly obtain the global optimal path and improve the convergence speed and accuracy.

关 键 词：自适应人工鱼群算法 清罐移动机器人 路径规划 最优路径 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]