English

Journal of Chinese Agricultural Mechanization

Journal of Chinese Agricultural Mechanization ›› 2025, Vol. 46 ›› Issue (4): 171-178.DOI: 10.13733/j.jcam.issn.2095-5553.2025.04.025

• Facilities Agriculture and Plant Protection Machinery Engineering • Previous Articles     Next Articles

Multi‑sensor fusion mapping and navigation research for tomato greenhouse robot

Fu Honglong1, 2, Hu Yubing1, 2, Xie Limin1, 2, Cai Yun1, 2, Fang Bing1, 2   

  1. (1. College of Mechanical and Electrical Engineering, Fujian Agriculture and Forestry University, Fuzhou, 350002, China;  2. Fujian Key Laboratory of Agricultural Information Sensing Technology, Fuzhou, 350002, China)
  • Online:2025-04-15 Published:2025-04-18

番茄温室机器人的多传感器融合建图与导航研究

傅泓龙1,2,胡裕兵1,2,谢立敏1,2,蔡云1,2,方兵1,2   

  1. (1. 福建农林大学机电工程学院,福州市,350002; 2. 福建省农业信息感知技术重点实验室,福州市,350002)
  • 基金资助:
    福建省林业科学技术攻关项目(2023KFJ01)

Abstract: According to the corridor environment and vegetation distribution in tomato greenhouse, SLAM and navigation algorithm of robot multi‑sensor fusion are optimized. Firstly, based on the Cartographer algorithm, the dedistortion processing of LiDAR data was carried out, and then the unscented Kalman filter (UKF) was used to fuse the information of LiDAR, odometer and IMU to optimize the robot pose estimation, and the optimized algorithm was used to construct a high‑precision map of the greenhouse. The established high‑precision raster map is verified by the improved A* and DWA fusion algorithm, and the search efficiency and path safety of the A* algorithm are improved by dynamically adjusting the weights and optimizing the search logic, so that the robot can find the optimal and safe path more intelligently in the greenhouse environment, and the effectiveness of the algorithm is verified in the gazebo platform of ROS and the greenhouse environment in the field. The experimental results indicated that the average position deviation was 10.3 cm when the running speed of the robot was not more than 0.6 m/s, which met the operation requirements of the tomato greenhouse.

Key words: tomato, greenhouse robot, multi?sensor fusion, laser SLAM, path planning, navigation

摘要: 针对番茄温室中的长廊环境及植被分布问题,对番茄机器人的多传感器融合的SLAM和导航算法进行优化。基于Cartographer算法,进行激光雷达数据去畸变处理,再通过无迹卡尔曼滤波(UKF)融合激光雷达、里程计及IMU信息进行机器人位姿估计优化,并使用优化后的算法对温室进行高精度地图构建。将建立的高精度栅格地图使用改进的A*与DWA融合算法进行验证,通过动态调整权重和优化搜索逻辑,提高A*算法的搜索效率和路径安全性,使得机器人在温室环境下能够更加智能地寻找最优且安全的路径,在ROS的gazebo平台和实地的温室环境中验证算法的有效性。结果表明,在机器人运行速度≤0.6 m/s时,平均位置偏差为10.3 cm,满足番茄温室的作业要求。

关键词: 番茄, 温室机器人, 多传感器融合, 激光SLAM, 路径规划, 导航

CLC Number: