狭窄环境中基于几何法的全局路径规划新方法  

A new global path planning method based on geometry algorithm in a narrow environment

作　　者：刘亚杰[1] 王航宇[1] 谢君[1] 

机构地区：[1]海军工程大学电子工程学院,武汉430033

出　　处：《海军工程大学学报》2010年第4期82-86,共5页Journal of Naval University of Engineering

摘　　要：研究狭窄障碍环境下基于几何法的移动机器人全局路规划方法。用不同多边形表示机器人和障碍物,多边形集合构成环境地图。利用数组矩阵存储机器人和障碍物的顶点坐标,便于计算机进行识别、分析和计算。在此基础上,建立了两个子函数——障碍物筛选子函数和凸包计算子函数。通过对两个子函数的循环调用,找出所有较优无碰路径,最后根据一定准则选择全局最优路径。该算法把狭窄障碍环境中的路径规划问题转换成凸包计算问题,且能够生成多条可供替换的较优路径。当环境空间相对狭窄、机器人形状较为复杂,在路径转弯处作旋转运动时,可根据安全需要选择合适的运动路径,从而增加了算法的适用性。仿真结果表明:该算法简便高效,能够满足路径实时规划要求。A global path planning method for a robot based on the geometry algorithm in a narrow working space with obstacles was analyzed.The robot and obstacles were expressed with different polygons,and the environment map was composed of gather of polygons.The vertex coordinates of robot and obstacles were stored by means of an array of matrix so as to be convenient for computer recognition,analysis and calculation.Two sub-functions of obstacles： selecting and convex hall calculating were established based on the environment modeling.All optimization collision free paths were found out with the two sub-functions transferred circularly.The optimization path was selected according to a certain rule finally.With this algorithm,the question of path planning in a narrow environment with obstacles is transformed to convex hall calculation,and more optimized and replaceable paths are provided.When the robot is more complicated in shape operating in a little narrow working space and turning on the path,an adaptable path can be selected according to safe requirements.The simulation results show the algorithm is so simple and highly efficient that it can meet real time application.

关 键 词：移动机器人 全局路径规划 几何法 凸包 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]