Analysis of Traveling Performance for Orchard Weeding Robot based on GNSS

Weeding is a necessary work for the growth and quality of crops, and large and small accidents occur or burden the workers during weeding. Therefore, to keep workers safe and reduce work burden, research on weeders is being conducted, and research on unmanned weeders capable of weeding without workers is also being conducted. In this study, we developed and tested a path-following algorithm that can drive along pre-recorded route points for unmanned driving and weeding of weeding robots. The path estimation algorithm used in this study determined the driving direction of the weeding robot based on the geometric model, which defined the steering angle as the angle between the global path and the straight line formed from the center coordinate point of the robot to the target waypoint. As a result of the driving test, the average driving error, maximum driving error, and maximum driving error width compared to the planned route were 0.1048m, 0.2341m, and 0.434m in the straight driving section, and 0.1279m, 0.4101m, and 0.5208m in the turning section.