Scholarlink

<p>Nowadays, autonomous ground vehicles (AGVs) transform urban transport, logistics, and field operations, and they occupy a strategic core role in autonomous driving commercialization.​ Although the artificial potential field method (APF) has made some achievements in trajectory planning of AGVs, it is obviously inappropriate to directly apply traditional methods to dynamic traffic scenarios. Vehicle dynamics environment, road regulations, and local minimum problems of artificial potential field methods should also be considered. Therefore, this paper introduces an efficient method for trajectory planning of autonomous ground vehicles using tangent-based artificial potential field (TAPF), which aims to address road dynamic obstacles and effectively resolve the local minimum problem. First, a structured potential field model is constructed, including the target gravitational field, obstacle repulsion field, and road boundary repulsion field. Additionally, in the trajectory generation process, the negative gradient method is employed to calculate the global potential field to generate a virtual resultant force to guide the vehicle’s movement. In the algorithm process, when the navigation line is detected to intersect with the obstacle safety circle, the candidate tangent points on the obstacle circle are generated and comprehensively considered, including geometric distance, historical path efficiency, and smoothness, and a tangential force vector is dynamically synthesized based on this. ​​ The tangential force integrates into the vehicle force analysis, directly breaks the force balance at local minima, guides the vehicle to bypass obstacles efficiently. Finally, simulation results across diverse traffic scenarios show that average lateral accelerations and max curvatures of planning trajectory with the proposed TAPF method are much smaller than those of classical APF by approximately 33.5% and 33.2%, while the increase of success rates for planning trajectory is about 40% compared with classical APF, which confirms that developed TAPF method can achieve real-time smooth, collision-free trajectory under high sampling frequency with 50&#xa0;Hz.</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

A Novel Trajectory Planning Strategy for Autonomous Ground Vehicles Using Tangent-based Artificial Potential Field

  • 期刊论文

  • 发表日期：

    2025年11月10日

查看全文
  • Xianjian Jin,
  • Jianning Lu,
  • Yinchen Tao,
  • Jianbo Lv,
  • Nonsly Valerienne,
  • Opinat Ikiela,
  • Guodong Yin

摘要

Nowadays, autonomous ground vehicles (AGVs) transform urban transport, logistics, and field operations, and they occupy a strategic core role in autonomous driving commercialization.​ Although the artificial potential field method (APF) has made some achievements in trajectory planning of AGVs, it is obviously inappropriate to directly apply traditional methods to dynamic traffic scenarios. Vehicle dynamics environment, road regulations, and local minimum problems of artificial potential field methods should also be considered. Therefore, this paper introduces an efficient method for trajectory planning of autonomous ground vehicles using tangent-based artificial potential field (TAPF), which aims to address road dynamic obstacles and effectively resolve the local minimum problem. First, a structured potential field model is constructed, including the target gravitational field, obstacle repulsion field, and road boundary repulsion field. Additionally, in the trajectory generation process, the negative gradient method is employed to calculate the global potential field to generate a virtual resultant force to guide the vehicle’s movement. In the algorithm process, when the navigation line is detected to intersect with the obstacle safety circle, the candidate tangent points on the obstacle circle are generated and comprehensively considered, including geometric distance, historical path efficiency, and smoothness, and a tangential force vector is dynamically synthesized based on this. ​​ The tangential force integrates into the vehicle force analysis, directly breaks the force balance at local minima, guides the vehicle to bypass obstacles efficiently. Finally, simulation results across diverse traffic scenarios show that average lateral accelerations and max curvatures of planning trajectory with the proposed TAPF method are much smaller than those of classical APF by approximately 33.5% and 33.2%, while the increase of success rates for planning trajectory is about 40% compared with classical APF, which confirms that developed TAPF method can achieve real-time smooth, collision-free trajectory under high sampling frequency with 50 Hz.