Scholarlink

<p>To address the issue that path planning algorithms in common indoor/outdoor obstacle environments fail to simultaneously maintain higher efficiency, shorter paths, and fewer turning points, a bidirectional scanning and sampling algorithm with virtual obstacles (BSS-VO) is proposed by integrating grid-based and sampling-based methods in this study. First, a bidirectional scanning and sampling method combining with virtual obstacles (BSS-VO) is proposed, which reduces computational load compared to grid-based algorithms while avoiding massive random and useless sampling typical of sampling-based approaches. Second, reverse search-based path optimization shortens generated path lengths and decreases turning points, thereby reducing motion time and enhancing trajectory smoothness for mobile robots. Third, simulation experiments were conducted in five typical simulation environments, with comparisons made against the Dijkstra, A*, D*, Jump Point Search (JPS), RRT, and RRT* algorithms. The simulation experiment results demonstrated that although the BSS-VO algorithm is slightly longer than the JPS algorithm in terms of planning time, the gap is within 0.1&#xa0;s. However, the BSS-VO algorithm has significant improvements compared with other algorithms in terms of path length and the number of turning points. Finally, physical experiments were conducted using mobile robots in a real-world environment comprising three classrooms connected by a shared corridor, with comparative analysis performed against the Dijkstra, A*, D*, Jump Point Search (JPS), RRT, and RRT* algorithms. The physical experimental results in real-world environments confirmed that the BSS-VO algorithm significantly outperforms the Dijkstra, A*, and D* algorithms in efficiency metrics. Path planning time was reduced by over 89%, while the number of turning points decreased by more than 40%. Compared to the Jump Point Search (JPS) algorithm, planning time increased by approximately 0.1s, but turning points were reduced by over 23%, which is more conducive to path planning and execution for mobile robots in physical settings. Additionally, dynamic environment experiments demonstrated that the BSS-VO algorithm exhibits enhanced adaptability to dynamic scenarios when integrated with local path planning methods.</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 path planning algorithm for mobile robots combining bidirectional scanning and sampling with virtual obstacles

  • 期刊论文

  • 发表日期：

    2025年12月19日

查看全文
  • Yi Wang,
  • Junyao Gao,
  • Weimeng Song,
  • Zihan Zhou,
  • Hongmei Zhao,
  • Juan Li,
  • Yufei Xu

摘要

To address the issue that path planning algorithms in common indoor/outdoor obstacle environments fail to simultaneously maintain higher efficiency, shorter paths, and fewer turning points, a bidirectional scanning and sampling algorithm with virtual obstacles (BSS-VO) is proposed by integrating grid-based and sampling-based methods in this study. First, a bidirectional scanning and sampling method combining with virtual obstacles (BSS-VO) is proposed, which reduces computational load compared to grid-based algorithms while avoiding massive random and useless sampling typical of sampling-based approaches. Second, reverse search-based path optimization shortens generated path lengths and decreases turning points, thereby reducing motion time and enhancing trajectory smoothness for mobile robots. Third, simulation experiments were conducted in five typical simulation environments, with comparisons made against the Dijkstra, A*, D*, Jump Point Search (JPS), RRT, and RRT* algorithms. The simulation experiment results demonstrated that although the BSS-VO algorithm is slightly longer than the JPS algorithm in terms of planning time, the gap is within 0.1 s. However, the BSS-VO algorithm has significant improvements compared with other algorithms in terms of path length and the number of turning points. Finally, physical experiments were conducted using mobile robots in a real-world environment comprising three classrooms connected by a shared corridor, with comparative analysis performed against the Dijkstra, A*, D*, Jump Point Search (JPS), RRT, and RRT* algorithms. The physical experimental results in real-world environments confirmed that the BSS-VO algorithm significantly outperforms the Dijkstra, A*, and D* algorithms in efficiency metrics. Path planning time was reduced by over 89%, while the number of turning points decreased by more than 40%. Compared to the Jump Point Search (JPS) algorithm, planning time increased by approximately 0.1s, but turning points were reduced by over 23%, which is more conducive to path planning and execution for mobile robots in physical settings. Additionally, dynamic environment experiments demonstrated that the BSS-VO algorithm exhibits enhanced adaptability to dynamic scenarios when integrated with local path planning methods.