自适应蚁群算法在轮式机器人三维路径规划中的应用  

The Application of Three-dimensional Path Planning of Wheeled Robot Based on Adaptive Ant Colony System Algorithm

作　　者：周利坤[1,2] 刘宏昭[1] 

机构地区：[1]西安理工大学机仪学院,西安710048 [2]武警工程大学军交运输系,西安710086

出　　处：《机械科学与技术》2013年第1期54-58,共5页Mechanical Science and Technology for Aerospace Engineering

基　　金：陕西省重点学科建设专项资金项目(102-00X903)资助

摘　　要：为了优化轮式机器人三维路径,进行了特殊三维空间有效路径设计,提出了自适应蚁群算法(AACS)。并将该算法应用于三维空间机器人路径规划中,将轮式机器人所处位置与目的点之间的空间划分成带有坡度角的立体网格,定义其有效路径,形成TSP模式。自适应蚁群按TSP模式搜索从原点到目的点之间的最短路径。实验表明:自适应蚁群优化方法克服了传统蚁群算法易陷于局部极值、搜索质量差和精度不高的缺点,提高了收敛速度和精度,输出稳定性好,可以解决轮式机器人在三维实际工作环境中的路径优化问题。To optimize three-dimensional path of wheeled robot, the special three-dimensional space-efficient path was designed, and adaptive ant colony system algorithm （AACS） was proposed to be applied to three-dimensional space path planning; the space between initial location point of wheeled robot and purpose point was divided into three-dimensional grid with a slope angle, and the effective path was defined to make a TSP model. Adaptive ant colony system algorithm searches the shortest path from initial point to destination point as TSP model. Experiments show that AACS overcomes the shortcoming of traditional ant colony algorithm of being easily trapped to local mini- ma, the search of poor quality and accuracy. AACS improves the convergence speed and accuracy, outputs stabili- ty, and solve three-dimensional path optimization problem for wheeled robots in practice.

关 键 词：机械学 轮式机器人 三维路径规划 自适应蚁群算法 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]