Path tracking control method for tracked agricultural vehicles based on slip-aware look-ahead point offset

Tracked agricultural vehicles operating in complex farmland environments are prone to track slip, which degrades path-tracking accuracy and may lead to unstable motion. To address the limitations of conventional geometric tracking algorithms under slip conditions, this study proposes a slip-aware look-ahead point offset path-tracking control method for tracked agricultural machinery.

An extended Kalman filter (EKF) is developed to fuse RTK-IMU pose measurements with track wheel-speed feedback, enabling real-time estimation of left and right track slip ratios. Based on the estimated slip difference, a target-point offset compensation mechanism is constructed, and the offset angle is optimized online using an improved particle swarm optimization (PSO) algorithm with a Chebyshev-window-based inertia weight strategy. In addition, a fuzzy controller is employed to adaptively adjust the look-ahead distance according to vehicle speed and path curvature, while a first-order low-pass filter is applied to smooth the commanded velocities.

Simulation results demonstrate that the proposed method significantly reduces lateral tracking errors and maintains smooth trajectories under severe slip conditions. Field experiments conducted at speeds of 0.35 m/s and 0.75 m/s show that the proposed method reduces the maximum lateral deviation by 78.1% and the average deviation by 50.6% compared with the traditional fuzzy pure pursuit algorithm. At 0.75 m/s, the maximum and average deviations are further reduced by 63.1% and 57.6%, respectively.

The results confirm that incorporating slip estimation and slip-aware target-point offset compensation effectively enhances path-tracking accuracy and robustness for tracked agricultural vehicles operating on soft and high-slip terrain. The proposed lightweight control framework provides a practical and reliable solution for autonomous navigation and plant-protection operations in complex farmland environments.