DESIGN AND TEST OF PATH TRACKING CONTROL SYSTEM FOR SOYBEAN WEEDING ROBOT

Abstract

To mitigate tracking degradation caused by unstable speeds in weeding robots, this study integrates Linear Active Disturbance Rejection Control (LADRC) with the Pure Pursuit (PP) algorithm. An Improved Northern Goshawk Optimization (INGO) algorithm is employed to optimize the LADRC parameters, enabling more precise speed regulation. Field experiments conducted at speeds of 0.5, 0.8, and 1.0 m/s compared the proposed approach with a conventional PID-PP controller. The results demonstrate that the proposed method reduced the maximum lateral tracking error by 9.67%, 19.0%, and 20.5%, respectively, while consistently improving both MAE and RMSE. These findings confirm that the proposed control strategy effectively enhances path tracking stability and precision, thereby improving the autonomous navigation performance of weeding robots.