T-S Fuzzy Tube Model Prediction Trajectory Tracking Control of Automatic Truck With Network Delay and Variable Load

The trajectory tracking performance is the foundation for self-driving trucks. To address the effects of communication delay and load variation on the trajectory tracking accuracy of automatic trucks, this article proposes a T-S fuzzy tube model prediction trajectory tracking control strategy. It includes the T-S fuzzy extended observer and the adaptive tube model prediction control. The T-S fuzzy extended observer reconstructs the fuzzy dynamic model of the automatic truck based on CAN occupancy rate (OR) and load, thereby reducing the deviation between the nominal model and the actual truck. Meanwhile, the T-S fuzzy extended observer derives the perturbation bounds of the random delay and tire cornering stiffness based on the CAN OR and load, respectively, to reduce the perturbation range of the uncertainty coefficients. The adaptive tube model prediction control reduces the conservatism of the controller by designing the shape of the tube and auxiliary control laws based on the fuzzy dynamics model and perturbation bounds. The experimental results show that the proposed method effectively reduces the trajectory tracking error of automatic trucks and improves the robustness of automatic trucks to changes in communication delay and load.