农业机械学报2025，Vol.56Issue(1)：123-132,10.DOI:10.6041/j.issn.1000-1298.2025.01.013
基于激光SLAM和AprilTag融合的温室移动机器人自主导航方法
Autonomous Navigation Method of Greenhouse Mobile Robot Based on Laser SLAM and AprilTag
摘要
Abstract
To enhance the precision and effectiveness of autonomous navigation for mobile robots in a greenhouse environment,a localization and autonomous navigation method was proposed by fusing laser simultaneous localization and mapping(SLAM)and AprilTag visual fiducial system.A mobile robot platform for greenhouse was developed based on multi-dimensional LiDAR,high-definition industrial cameras,and edge computing devices.Firstly,in order to improve the efficiency and accuracy of map building,Gmapping algorithm was used to construct a two-dimension grid map by taking two-dimension LiDAR point cloud processed by three-dimension LiDAR point cloud and the odometer data obtained by RF2O algorithm as inputs.Then AprilTag positioning correction method was proposed to solve the positioning loss problem of the mobile robot for the environmental characteristics of narrow,symmetrical,and repetitive facility cultivation ridges.Finally,a combination of Dijkstra algorithm and dynamic window approach(DWA)algorithm was used to plan the global and local navigation paths.The positioning accuracy of mobile robot was evaluated based on a three-dimensional motion capture system in laboratory.The experimental results showed that at speeds of 0.4 m/s,0.3 m/s and 0.2 m/s,the average longitudinal positioning error of the mobile robot was no more than 0.066 m,and the standard deviation was no more than 0.049 m.The average lateral positioning error was no more than 0.117 m,and the standard deviation was no more than 0.092 m.Autonomous navigation performance evaluation tests of mobile robot were carried out in a greenhouse environment.The experimental results showed that at speeds of 0.4 m/s,0.3 m/s and 0.2 m/s,the average lateral deviation between the actual driving trajectory of the mobile robot and the expected trajectory was no more than 0.050 m,the standard deviation was no more than 0.032 m,and the average heading deviation was no more than 2.2°,the standard deviation was no more than 1.4°.The positioning and navigation accuracy of mobile robots can meet the continuous and stable operation requirements in greenhouse.关键词
温室/移动机器人/多传感器融合/自主导航/定位校正/路径规划Key words
greenhouse/mobile robot/multi-sensor fusion/autonomous navigation/positioning correction/path planning分类
农业科技引用本文复制引用
张文翔,卢鑫羽,张兵园,贡宇,任妮,张美娜..基于激光SLAM和AprilTag融合的温室移动机器人自主导航方法[J].农业机械学报,2025,56(1):123-132,10.基金项目
江苏省农业科技自主创新资金项目(CX(22)5009) （CX(22）
