Scholarlink

<p>Model predictive control (MPC) without guaranteed stability is typically employed for trajectory tracking of autonomous trucks (ATs). However, in certain scenarios, the tracking error may fail to converge. To address this, the optimization control problem can be designed by incorporating terminal ingredients, i.e., the terminal control gain, terminal constraint, and terminal cost function. In this paper, we propose a nonlinear model predictive control (NMPC) scheme with terminal ingredients for trajectory tracking of ATs in the presence of the coupled longitudinal and lateral dynamics. The trajectory tracking problem exhibits asymptotic convergence, and the optimization control problem (OCP) ensures recursive feasibility. The complexity of the proposed controller is similar to that of a standard NMPC without terminal ingredients. Additionally, we introduce an efficient Newton-type method with a look-up table (NTLT) to solve the OCP. Co-simulations in Matlab/Simulink and TruckSim validate the effectiveness of the proposed NMPC scheme and the NTLT across various scenarios.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Nonlinear model predictive control of trajectory tracking for autonomous trucks with terminal ingredients

  • 期刊论文

  • 发表日期：

    2025年5月9日

查看全文
  • Wenbo Li,
  • Shuyou Yu,
  • Baojun Lin,
  • Hong Chen

摘要

Model predictive control (MPC) without guaranteed stability is typically employed for trajectory tracking of autonomous trucks (ATs). However, in certain scenarios, the tracking error may fail to converge. To address this, the optimization control problem can be designed by incorporating terminal ingredients, i.e., the terminal control gain, terminal constraint, and terminal cost function. In this paper, we propose a nonlinear model predictive control (NMPC) scheme with terminal ingredients for trajectory tracking of ATs in the presence of the coupled longitudinal and lateral dynamics. The trajectory tracking problem exhibits asymptotic convergence, and the optimization control problem (OCP) ensures recursive feasibility. The complexity of the proposed controller is similar to that of a standard NMPC without terminal ingredients. Additionally, we introduce an efficient Newton-type method with a look-up table (NTLT) to solve the OCP. Co-simulations in Matlab/Simulink and TruckSim validate the effectiveness of the proposed NMPC scheme and the NTLT across various scenarios.