Scholarlink

<p>This paper proposes a novel time-efficient hierarchical vehicle control framework (THVCF) to meet the real-time and high-accuracy requirements of intelligent vehicle path tracking. The main contribution lies in the integration of three coordinated layers: an offline explicit control layer using EMPC to precompute control laws and reduce online computational cost; an online control layer combining feed-forward and feedback strategies to improve steering accuracy; and a parameter optimization layer that adaptively tunes control weights for different paths. This layered design enhances tracking performance, robustness, and real-time capability. Simulation results under double lane change and circular arc scenarios demonstrate that THVCF improves tracking accuracy and computational efficiency by over 60% and 80%, respectively. Furthermore, steering overshoot under high-curvature conditions is reduced by 56% through feed-forward compensation and adaptive tuning. Ablation studies further verify the contribution of each layer. These results confirm that THVCF offers a practical and scalable solution for real-time vehicle control in complex driving environments.</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

Time-Efficient Hierarchical Vehicle Control Framework for Path Tracking in Intelligent Vehicles

  • 期刊论文

  • 发表日期：

    2025年5月26日

查看全文
  • Ming Liu,
  • Yuanjian Zhang,
  • Bo Leng,
  • Jincheng Hu,
  • Jihao Li,
  • Lu Xiong

摘要

This paper proposes a novel time-efficient hierarchical vehicle control framework (THVCF) to meet the real-time and high-accuracy requirements of intelligent vehicle path tracking. The main contribution lies in the integration of three coordinated layers: an offline explicit control layer using EMPC to precompute control laws and reduce online computational cost; an online control layer combining feed-forward and feedback strategies to improve steering accuracy; and a parameter optimization layer that adaptively tunes control weights for different paths. This layered design enhances tracking performance, robustness, and real-time capability. Simulation results under double lane change and circular arc scenarios demonstrate that THVCF improves tracking accuracy and computational efficiency by over 60% and 80%, respectively. Furthermore, steering overshoot under high-curvature conditions is reduced by 56% through feed-forward compensation and adaptive tuning. Ablation studies further verify the contribution of each layer. These results confirm that THVCF offers a practical and scalable solution for real-time vehicle control in complex driving environments.