Orchard mobile robot and headland turning method

Abstract

The invention relates to a headland turning method for an orchard mobile robot. The method includes the following steps that: step 1, a Faster R-cnn neural network performs trunk recognition; step 2,the end position of a fruit tree row is identified on the basis of trunk recognition; step 3, after the robot reaches the end of the row, a turning radius d is determined through a binocular camera; step 4, a controller controls wheels to rotate so as to start turning, an actual turning radius is compared with the determined turning radius d, the turning posture of the robot is adjusted in time; and step 5, when a turning angle measured by an electronic compass reaches 180 degrees, headland turning is completed.

Description

technical field [0001] The invention belongs to the technical field of robot steering, and in particular relates to an orchard mobile robot and a ground steering method. Background technique [0002] At present, the orchard is developing in the direction of automation, informatization, and intelligence, in order to reduce labor intensity, save labor, improve labor efficiency, and realize fine operation of the orchard. Orchard robots that can automatically complete operations such as inspection, pruning, spraying, and picking continue to enter various orchards. Orchard robots need a series of basic functions to complete operations, such as autonomous navigation technology for orchard robots. At present, the research focus on the autonomous navigation technology of orchard robots is mostly on the positioning and navigation within a certain fruit tree row, and the research on the end-of-row positioning and ground turning is insufficient. [0003] At present, there are few meth...

AI Technical Summary

Problems solved by technology

[0004] For row-end positioning, the method of using lidar to set reflective tape at the end of the row needs to install reflective tape at the end of each fruit tree row, which is time-consuming, labor-intensive, and costly; in the orchard environment, the branches and leaves are dense, GPS signals will be blocked, and the positioning is not accurate.

For headland steering, there is no reference object when the angle is changed every time to reach a new straight line by the method of straight instead of curved, which will make the error accumulate; the method of right-angle turning cannot guarantee that the body is just parallel to the fruit tree row before the first rotation, and the first time If the angle deviation between the straight line after turning and the vertical line of the fruit tree row is too large, it will not be conducive to the smooth entry into the new fruit tree row after the second rotation of the body; Entering the adjacent fruit tree row can only cross the adjacent row and enter the next row of the adjacent row

Images