Abstract
In this study, a wheel-type robot tractor was developed in order to reduce agricultural labor. An RTK-GPS (real-time kinematic global positioning system) and an IMU (inertial measurement unit) were used as position and attitude sensors for the navigation system. A path-following navigation algorithm and an algorithm for turning at headland were designed in this study. Tuning of the auto-steering system to improve the navigation accuracy was done. Applications of the robot tractor including tillage, spraying and weeding were tested in real farms. The RMS (root mean square) of lateral error was less than 0.05 m for all of the applications. The results of the tests revealed that the robot tractor has high and stable navigation accuracy for the tested operations.
