Scholarlink

<p>A single-actuator unmanned aerial vehicle (UAV) named PULSAR II is driven by one motor and achieves three-dimensional maneuverability and natural self-rotation. This motion enables an omnidirectional field of view (FoV) for light detection and ranging (LiDAR), thereby eliminating unobserved areas. To enhance agility, we first built an accurate propulsion model using a data-driven method and compensated its nonlinearities. Then, we designed model predictive controller on the <i>S</i><sup>2</sup> manifold for minimal parameterization and aggressive tracking performance. We also derived the UAV’s differential flatness to simplify trajectory optimization. To ensure collision resistance, we designed a protective solid frame structure around the UAV and developed a disturbance observer to reject collision-induced disturbances in a fictional coordinate frame that decouples self-rotation, with parameters optimized by <i>μ</i> synthesis to address model uncertainty. By integrating all modules, we established an autonomous UAV system that demonstrated agile, robust, and fully autonomous flight in real-world 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

An autonomous single-actuator UAV with omnidirectional field of view, high agility, and collision resistance

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月21日

查看全文
  • Nan Chen,
  • Haotian Li,
  • Yunfan Ren,
  • Guozheng Lu,
  • Fangcheng Zhu,
  • Jiarong Lin,
  • Fu Zhang

摘要

A single-actuator unmanned aerial vehicle (UAV) named PULSAR II is driven by one motor and achieves three-dimensional maneuverability and natural self-rotation. This motion enables an omnidirectional field of view (FoV) for light detection and ranging (LiDAR), thereby eliminating unobserved areas. To enhance agility, we first built an accurate propulsion model using a data-driven method and compensated its nonlinearities. Then, we designed model predictive controller on the S2 manifold for minimal parameterization and aggressive tracking performance. We also derived the UAV’s differential flatness to simplify trajectory optimization. To ensure collision resistance, we designed a protective solid frame structure around the UAV and developed a disturbance observer to reject collision-induced disturbances in a fictional coordinate frame that decouples self-rotation, with parameters optimized by μ synthesis to address model uncertainty. By integrating all modules, we established an autonomous UAV system that demonstrated agile, robust, and fully autonomous flight in real-world environments.