In this paper, we focus on the integration of motion planning and control for high-performance automated vehicles. We propose a novel tube-based nonlinear model predictive control (TNMPC) scheme that combines the strengths of MPC and boundary layer sliding control (BLSC) to drive the controlled vehicle along a real-time and high-performance trajectory while simultaneously ensuring practical satisfaction of constraints. Specifically, our approach uses an extended kinematic vehicle model with slip angle parameters to provide explicit expressions for modeling error, thereby facilitating the design of BLSC gains. We then reveal an inequality relation between the slip angles and acceleration, which allows the BLSC gains to be represented as functions of the optimization variables. This innovative approach enables the inclusion of constraints on the BLSC gains into the TNMPC motion planning problem, striking a balance among performance, modeling error, and practical control capability. The feasibility of the kinematic model is also guaranteed without the need for the commonly adopted acceleration constraint. Moreover, tightening of the state and input constraints in TNMPC is achieved through coordinate transformation of the tubes, interval analysis, and robust optimization theory. Simulations in racing scenarios demonstrate the robustness and effectiveness of the proposed TNMPC scheme, illustrating its ease of adaptability across various road friction conditions.
Integration of Motion Planning and Control for High-Performance Automated Vehicles Using Tube-Based Nonlinear MPC
IEEE Transactions on Intelligent Vehicles ; 9 , 2 ; 3859-3875
2024-02-01
2867492 byte
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Probabilistic Motion Planning for Automated Vehicles
Katalog Medizin | 2021
|Probabilistic Motion Planning for Automated Vehicles
GWLB - Gottfried Wilhelm Leibniz Bibliothek | 2021
|Probabilistic Motion Planning for Automated Vehicles
GWLB - Gottfried Wilhelm Leibniz Bibliothek | 2021
|