A Vehicle Trajectory Tracking Method With a Time-Varying Model Based on the Model Predictive Control

• Main Authors: Xiaohua Song, Yiming Shao, Zhihua Qu
• Format: Article
• Language: English
• Published: IEEE 2020-01-01
• Series: IEEE Access
• Subjects: Model predictive control; stability; time-varying model control; vehicle trajectory tracking method
• Online Access: https://ieeexplore.ieee.org/document/8946576/

The vehicle trajectory tracking algorithm is one of the key and difficult issues of intelligent driving technologies. In current control algorithms for the vehicle trajectory tracking, there are three main assumptions: the standard working condition for the driving path, the same control model used for the entire control process, and a fixed value for the longitudinal vehicle speed. However, the above determinations in current control algorithms are inconsistent with the actual vehicle driving conditions. To overcome those problems, a vehicle trajectory tracking method with a time-varying model is proposed. The time-varying model is developed by using a two-dimensional vehicle kinematics model. This method considers the influences of the longitudinal speed and road curvature on the vehicle trajectory tracking stability under the low-speed complex driving condition. Thus, the proposed method can improve the trajectory tracking accuracy when the unmanned vehicle is located at the road with a sharp curvature under the low-speed complex driving condition. Moreover, the proposed model can achieve the real time calculation. Meanwhile, the prediction accuracy of the vehicle kinematics model is ensured. The proposed approach with the above characteristics can complete the trajectory tracking for the route composed of arbitrary curves. The results show that the proposed method can effectively improve the trajectory tracking stability of the unmanned vehicle on the roads with different curvatures under complex driving conditions.