Scholarlink

This paper presents a safe motion planning and control algorithm for four-wheeled autonomous vehicles in unstructured environments. First, a low-collision-risk, low-turning and straight-line dominant path is generated quickly for navigation, considering the massive irregularly placed obstacles in the unstructured environment. Then, a smooth, safe, velocity-controllable trajectory, which adheres to the vehicle’s nonholonomic constraints, is generated through sampling within the Frenet frame, also addressing the challenges posed by dynamic local obstacles. Finally, model predictive control is applied to ensure stable trajectory tracking. The proposed method is validated using a four-wheeled vehicle model within an unstructured environment developed in Unreal Engine, demonstrating enhanced safety performance.

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

Motion Planning and Control for Safe Driving of Autonomous Vehicle in Complex Unstructured Environments

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Zhenhua Zhang,
  • Jiang Bian,
  • Tianchang Nie,
  • Yuhua Li,
  • Xuezhi Liang,
  • Pengfei Guo,
  • boshan che,
  • Junhui Li,
  • Chaomeng Guo,
  • Menghan Jiao,
  • Haijie Guan

摘要

This paper presents a safe motion planning and control algorithm for four-wheeled autonomous vehicles in unstructured environments. First, a low-collision-risk, low-turning and straight-line dominant path is generated quickly for navigation, considering the massive irregularly placed obstacles in the unstructured environment. Then, a smooth, safe, velocity-controllable trajectory, which adheres to the vehicle’s nonholonomic constraints, is generated through sampling within the Frenet frame, also addressing the challenges posed by dynamic local obstacles. Finally, model predictive control is applied to ensure stable trajectory tracking. The proposed method is validated using a four-wheeled vehicle model within an unstructured environment developed in Unreal Engine, demonstrating enhanced safety performance.