Robust Model Predictive Control for Path Tracking of a Tracked Vehicle with a Steerable Trailer in the Presence of Slip

In comparison to general autonomous robots, high precision guidance of farm vehicles becomes more complex and challenging as these vehicles are inevitably subjected to significant slip caused by very rough and unsteady terrain which is sometimes undulating. Therefore, it is essential to design controllers that have the capability to react to significant and uncertain disturbances. This paper presents a robust model predictive controller for path tracking of a tracked vehicle towing a steerable trailer in the presence of unknown but bounded disturbances. Based on kinematics and a virtual spring, a virtual error vector model with slip is used to model path offset and orientation offset for the tracked vehicle and the trailer. An adaptive min-max model predictive control method is proposed to guarantee robustness and accuracy in the path tracking. Finally, the proposed controller is compared with a min-max model predictive controller and a sliding mode controller in a realistic dynamic simulation platform. The results prove that the proposed adaptive min-max model predictive controller provides the required accuracy and robustness in the presence of slip.