系统工程与电子技术 ›› 2025, Vol. 47 ›› Issue (4): 1206-1213.doi: 10.12305/j.issn.1001-506X.2025.04.17

• 系统工程 • 上一篇    下一篇

面向飞机表面视觉检查的无人机覆盖路径规划

陈威1, 王从庆1,*, 曾强2, 李战2   

  1. 1. 南京航空航天大学自动化学院, 江苏 南京 211106
    2. 中国航空工业集团有限公司贵州飞机有限责任公司, 贵州 安顺 561000
  • 收稿日期:2023-08-28 出版日期:2025-04-25 发布日期:2025-05-28
  • 通讯作者: 王从庆
  • 作者简介:陈威 (1998—), 男, 硕士, 主要研究方向为模式识别与智能系统
    王从庆 (1960—), 男, 教授, 博士, 主要研究方向为模式识别与智能系统
    曾强 (1987—), 男, 高级工程师, 硕士, 主要研究方向为无损检测
    李战 (1969—), 男, 高级工程师, 博士, 主要研究方向为计量检测
  • 基金资助:
    国家自然科学基金(61573185);“慧眼行动”成果转化项目(62502010224)

UAV coverage path planning for aircraft surface visual inspection

Wei CHEN1, Congqing WANG1,*, Qiang ZENG2, Zhan LI2   

  1. 1. College of Automation, Nanjing University of Aeronautics and Astronautics, Nanjing 211106, China
    2. Aviation Industry Corporation of China(Guizhou)Company, Limited, Aviation Industry Corporation of China, Ltd., Anshun 561000, China
  • Received:2023-08-28 Online:2025-04-25 Published:2025-05-28
  • Contact: Congqing WANG

摘要:

为了高效规划无人机执行飞机表面视觉检查任务时的飞行路径, 提出一种基于自适应混合采样策略的覆盖路径规划算法, 通过视点生成、视点筛选和覆盖路径规划求解无人机最优检查路径。首先, 基于待检查飞机模型进行视线最优采样和自适应补充采样, 生成冗余视点集合。然后, 采用一种基于动态加权启发式的图搜索算法, 搜索并选择一组提供增量覆盖的有效视点。最后, 在原Lin-Kernighan启发式(Lin-Kernighan heuristic, LKH)算法中设计了路径碰撞检测模块, 并通过改进后的LKH算法求解无人机无碰撞检查路径。仿真实验结果表明, 所提算法在两种不同场景下规划出的无人机检查路径最大飞机表面覆盖率分别为93.44%和96.44%, 在路径长度、视点数量和算法耗费时间方面均优于其他对比算法。

关键词: 飞机表面检查, 覆盖路径规划, 自适应混合采样, 图搜索算法, 无人机

Abstract:

In order to efficiently plan the flight paths for unmanned aerial vehicle (UAV) conducting visual inspection tasks on aircraft surface, a coverage path planning algorithm is proposed by using an adaptive hybrid sampling strategy, which involves viewpoint generation, viewpoint selection, and coverage path planning to determine the optimal inspection path for UAV. Firstly, optimal line-of-sight sampling and adaptive supplementary sampling are performed based on the aircraft model to be inspected to generate a redundant set of viewpoints. Secondly, a heuristic graph search algorithm based on dynamic weighting is adopted to search and select a set of effective viewpoints to offer incremental coverage. Finally, a path collision detection module is designed in the original Lin-Kernighan heuristic (LKH) algorithm and the improved LKH algorithm is used to solve UAV collisionless inspection path.Simulation experimental results demonstrate that the proposed method achieves max aircraft surface coverage rates of 93.44% and 96.44% of the aircraft inspection paths planned in two different scenarios, which outperforms other comparative algorithms in terms of path length, number of viewpoints, and algorithmic time consumption.

Key words: aircraft surface inspection, coverage path planning, adaptive hybrid sampling, graph search algorithm, unmanned aerial vehicle (UAV)

中图分类号: