基于不确定网格地图的移动机器人导航

Navigation for mobile robot based on uncertainty grid-map

中文关键词  占据网格  模糊集  路径规划  传感器融合

英文关键词  occupancy grid  fuzzy sets  path planning  sensor fusion

中文摘要

      研究了在未知环境下的移动机器人导航问题. 在分析超声传感器不确定性模型的基础上, 根据模糊集理论创建网格地图来描述机器人工作环境, 使用模糊隶属度表示网格占用状态. 通过网格信息融合来减弱传感器测量误差, 提高网格地图的精度. 提出基于模糊网格地图的路径规划算法, 利用重复局部优化路径搜索来实现全局路径规划. 机器人通过交替进行创建地图和路径规划两个基本过程来完成导航任务. 仿真结果表明创建的地图能较精确地表示环境信息, 规划的路径可以使机器人安全地到达目的地.

英文摘要

      The problem of the mobile robot navigation is studied in this paper. Based on the uncertain model of ultrasonic sensor, the grid map is built according to fuzzy sets, which is utilized to describe the environments. The fuzzy membership is used to denote the state of occupancy grid. The measure error of sensor can be effectively reduced through fusing information of grids. Accordingly, the precision of map building is improved. Based on the fuzzy grid map, an approach of optimal path planning is also presented. Furthermore, local path searching is iterated, and is used to substitute the global path planning. The robot alternately executes two processes: map building and path planning in order to achieve navigation mission. Finally, simulation results show that the built map can accurately depict environment and the robot can successfully achieve the destination in the light of planned path.