2.5 Integrated Framework of Improved LeGO-LOAM and Enhanced OpenPlanner
In previous sections, a unique hardware system tailored for the greenhouse agricultural environment and enhanced localization and navigation algorithms were proposed. In this section, we will introduce the overall framework of autonomous navigation for greenhouse robots based on these innovations. As shown in Fig. 2, the entire system can be divided into two components: SLAM and navigation.
The implementation of SLAM is based on the Improved LeGO-LOAM. The contribution of this paper includes the integration of a relocalization module and the fusion of information from multiple sensors. These enhancements help to achieve higher accuracy in pose and relocalization information, as well as a high-quality 3D point cloud prior map. All of these contribute to improving the accuracy of the system’s information acquisition.
The navigation process consists of four modules, as detailed in Fig. 2. Initially, a global trajectory is formulated using a dynamic programming algorithm, dependent on the prior map and waypoints. Following this, modules dedicated to environmental perception, local planning, and trajectory tracking persistently execute until the designated target is achieved. Of importance is the enhancement of the local planner, contributing to an increased robustness in greenhouse navigation.
Table 2. Key Parameters of Experiment
Parameter Value
Lidar range 0.15m-150m
Lidar frequency 10HZ
LiDAR Scan Lines 16
Traversable Height 0.6m
Mini speed of dynamic obstacle 0.5 m/s