复杂低空环境下无人机自主定位技术研究进展
作者:
作者单位:

1西北工业大学无人系统技术研究院,西安710072;2西北工业大学人工智能学院,西安710072;3无人飞行器技术全国重点实验室,西安710072

作者简介:

通讯作者:

基金项目:

国家自然科学基金(42504030);中央高校基本科研业务费(科技专项)(D5000250047)。


A Review of Autonomous Localization Technologies for Unmanned Aerial Vehicles in Complex Low-Altitude Environments
Author:
Affiliation:

1Unmanned Systems Research Institute, Northwestern Polytechnical University, Xi’an 710072, China;2School of Artificial Intelligence, Northwestern Polytechnical University, Xi’an 710072, China;3National Key Laboratory of Unmanned Aerial Vehicle Technology, Xi’an 710072, China

Fund Project:

National Natural Science Foundation of China (No.42504030); Fundamental Research Fund for the Central Universities (Science and Technology Program) (No.D5000250047).

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
    摘要:

    复杂低空环境通常呈现出多源干扰叠加、感知条件剧烈变化与信息不完备并存等特征,对无人机自主定位的连续性、可靠性与可信性提出了严峻挑战。在此类环境下,全球卫星导航系统(Global navigation satellite system, GNSS)信号易受遮挡与干扰而失效,视觉感知面临弱纹理、强动态与光照突变等退化问题,惯性测量则不可避免地产生长期累积漂移,三者耦合作用显著削弱了定位系统的稳定性与鲁棒性。为此,本文系统梳理了低空典型退化环境类型,重点分析了多源混合干扰场景下视觉特征缺失、IMU误差发散与卫星定位性能退化等关键技术瓶颈。在此基础上,综述了无人机视觉导航定位技术的发展脉络,涵盖基于卫星/先验地图的视觉匹配定位方法以及视觉SLAM的最新研究进展;进一步总结了视觉-惯性系统融合建模与感知增强方法,阐明其在提升定位精度与稳健性方面的技术优势。随后,论述了多源融合导航框架及面向拒止环境的鲁棒融合策略,重点关注视觉、惯性、激光雷达以及卫星等多模态信息的协同建模、退化感知与完好性监测。最后,展望了数据驱动的多模态自适应导航方法以及轻量化、智能化的无人机高可信导航技术发展趋势。旨在为复杂低空环境下无人机高可靠自主定位技术的研究与工程应用提供系统参考。

    Abstract:

    Complex low-altitude environments are typically characterized by the superposition of multi-source interference, drastic variations in sensing conditions, and incomplete environmental information, which collectively pose significant challenges to the continuity, reliability, and integrity of autonomous localization for unmanned aerial vehicles (UAVs). In such scenarios, Global Navigation Satellite System (GNSS) signals are prone to blockage and interference, visual perception suffers from weak textures, dynamic disturbances, and abrupt illumination changes, and inertial measurements inevitably accumulate long-term drift. The coupled degradation of these sensing modalities substantially undermines the stability and robustness of localization systems. To address these challenges, this paper systematically reviews representative types of degraded low-altitude environments and analyzes key technical bottlenecks under multi-source hybrid interference, including visual feature loss, inertial error divergence, and satellite positioning performance deterioration. Building upon this analysis, the developmental trajectory of vision-based navigation and localization techniques for UAVs is comprehensively surveyed, covering visual matching methods based on satellite signals or prior maps as well as recent advances in visual simultaneous localization and mapping (SLAM). Furthermore, visual-inertial fusion modeling and perception enhancement strategies are summarized, highlighting their technical advantages in improving localization accuracy and robustness. Subsequently, multi-sensor fusion navigation frameworks and robust fusion strategies tailored for GNSS-denied or degraded environments are discussed, with particular emphasis on collaborative modeling, degradation awareness, and integrity monitoring across heterogeneous modalities, including vision, inertial sensors, LiDAR, and satellite positioning. Finally, the paper outlines future directions for data-driven multimodal adaptive navigation methods, as well as the development trends of lightweight and intelligent high-integrity navigation technologies for unmanned aerial vehicles. This survey aims to provide a systematic reference for the research and engineering implementation of highly reliable autonomous localization technologies for UAVs operating in complex low-altitude environments.

    参考文献
    相似文献
    引证文献
引用本文

许悦雷,王铉彬,薛尚捷,徐金海.复杂低空环境下无人机自主定位技术研究进展[J].数据采集与处理,2026,(2):592-619

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
历史
  • 收稿日期:2026-01-12
  • 最后修改日期:2026-02-28
  • 录用日期:
  • 在线发布日期: 2026-04-15