Path Tracking of Orchard Tractor Based on Linear Time-varying Model Predictive Control

The development of unmanned agricultural vehicle is of great significance for improving the level of agriculture in China. In this paper, a path tracking controller for unmanned tractors was designed based on linear time-varying Model Predictive Control (MPC). The objective is to improve orchard tractor path tracking accuracy and ensure high robustness and speed adaptability of the controller. Simulation was performed in the simulink-carsim joint platform by setting an orchard path. Because the tractor generally travels on a soft pavement, model parameters were modified by more than 10% to evaluate the robustness of the controller. The results showed that the designed controller is satisfactory, which can almost completely track the straight line. The mean lateral error for the curved lines is 7 cm at 2.5 m/s. Meanwhile, the controller has strong robustness and speed adaptability.