复杂场景下基于深度学习与多传感器融合的无人机配网巡检智能避障技术

doi:10.13474/j.cnki.11-2246.2025.0105

摘要/Abstract

摘要： 在电力配网的巡检过程中,复杂的环境条件,如树木遮挡和随机性障碍物,常常导致无人机在执行任务时遇到悬停、撞机等问题,严重影响巡检效率和安全性。为应对这一挑战,本文提出了一种针对复杂场景下的无人机自动巡检智能避障技术,开发了融合激光雷达和机器视觉的环境感知系统,通过利用空洞空间金字塔池化结构增大卷积核的感受野,捕获多尺度信息对障碍物进行实时识别,并利用先进的路径规划算法动态调整无人机的飞行路径,以避开障碍物。仿真测试验证表明,该系统在复杂环境中的避障能力得到显著提高,巡检效率提升了20%以上,且有效降低了事故风险。本文所提出的智能避障技术为电力配网的无人机巡检提供了一种高效、安全的解决方案,具备广泛的应用价值和推广前景。

关键词: 无人机, 空洞空间金字塔池化, 避障能力, 路径规划

Abstract: In the inspection process of electric power distribution networks, complex environmental conditions, such as tree shading and random obstacles, often lead to UAVs encountering problems such as hovering and crashing when performing their tasks, which seriously affects inspection efficiency and safety. To cope with this challenge, this paper proposes an intelligent obstacle avoidance technique for automatic UAV inspection in complex scenarios. An environmental sensing system fusing LiDAR and machine vision is developed to capture multi-scale information for real-time obstacle identification by utilizing an atrous spatial pyramid pooling structure to increase the sensory field of convolution kernel. Advanced path planning algorithms are utilized to dynamically adjust the UAV's flight path to avoid obstacles. The results on the simulation tests show that the system's obstacle avoidance ability in complex environments is significantly improved, the inspection efficiency is increased by more than 20%, and the risk of accidents is effectively reduced. The proposed intelligent obstacle avoidance technology provides an efficient and safe solution for UAV inspection of power distribution networks, which has a wide range of application value and promotion prospects.

Key words: unmanned aerial vehicle, atrous spatial pyramid pooling, obstacle avoidance, path planning

中图分类号:

P237

廖红兵, 况松陵, 李扬帆, 黄晓露, 王刚, 魏洪. 复杂场景下基于深度学习与多传感器融合的无人机配网巡检智能避障技术[J]. 测绘通报, 2025, 0(1): 22-28.

LIAO Hongbing, KUANG Songling, LI Yangfan, HUANG Xiaolu, WANG Gang, WEI Hong. Intelligent obstacle avoidance technology for UAV distribution network inspection based on deep learning and multi-sensor fusion in complex scenarios[J]. Bulletin of Surveying and Mapping, 2025, 0(1): 22-28.

参考文献

[1] 王果,胡青峰,龚天浩,等.倾斜摄影模型交互式建筑自动成图方法[J].测绘通报,2021(S1):308-311.
[2] 李春瑾,黄巧霞.基于经纬M300 RTK无人机在水文行业的应用[J].内蒙古水利,2023(5):33-34.
[3] 马逢.面向灭火救援的多旋翼无人机架构设计现状及改进探索[J].中国应急救援,2023(3):50-53.
[4] 冯欧阳,马奇,移强平,等.基于改进蚁群算法的无人机在配电网线路巡检中的研究与应用[J].电气时代,2024(7):53-57.
[5] 李鑫龙,李伟明,张清民.天狼星无人机摄影测量系统在大比例尺测图与三维建模中的应用[J].测绘通报,2018(10):159-162.
[6] 梁桂明,张玉姣.新常态下开展无人机倾斜实景三维生产的关键技术研究[J].测绘通报,2024(S2):175-177.
[7] 郭涵,康冰怡,王艺璇,等.基于视觉感知的无人机自主巡检系统设计与实现[J].中国高新科技,2024(7):105-107.
[8] LEONG W L,WANG Pengfei,HUANG Sunan,et al.Vision-based sense and avoid with monocular vision and real-time object detection for UAVs[C]//Proceedings of 2021 International Conference on Unmanned Aircraft Systems (ICUAS).Athens:IEEE,2021:1345-1354.
[9] ZHANG Zhouyu,ZHANG Youmin,CAO Yunfeng,et al.Infrared and visible airborne targets image fusion with applications to sense and avoid[J].IFAC-PapersOnLine,2020,53(2):14742-14747.
[10] DU Hao,WANG Wei,XU Chaowen,et al.Real-time onboard 3D state estimation of an unmanned aerial vehicle in multi-environments using multi-sensor data fusion[J].Sensors,2020,20(3):919.
[11] 贾秉健,刘博文,张润贤,等.配电网线路无人机智能巡检技术研究[J].信息记录材料,2024,25(1):130-132.
[12] 袁铭均.浅析基于多旋翼无人机的机载激光点云树障隐患的识别[J].中国设备工程,2024(18):40-42.
[13] AGGARWAL S,KUMAR N.Path planning techniques for unmanned aerial vehicles:a review,solutions,and challenges[J].Computer Communications,2020,149:270-299.
[14] ZHAI Yongjie,CHEN Rui,YANG Qiang,et al.Insulator fault detection based on spatial morphological features of aerial images[J].IEEE Access,2018,6:35316-35326.
[15] YAN Tiantian,YANG Guodong,YU Junzhi.Feature fusion based insulator detection for aerial inspection[C]//Proceedings of 36th Chinese Control Conference (CCC).Dalian:IEEE,2017:10972-10977.
[16] LOWE D G.Distinctive image features from scale-invariant keypoints[J].International Journal of Computer Vision,2004,60(2):91-110.
[17] DALAL N,TRIGGS B.Histograms of oriented gradients for human detection[C]//Proceedings of 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR '05).San Diego:IEEE,2005:886-893.
[18] BAY H,ESS A,TUYTELAARS T,et al.Speeded-up robust features (SURF)[J].Computer Vision and Image Understanding,2008,110(3):346-359.
[19] GIRSHICK R,DONAHUE J,DARRELL T,et al.Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of 2014 IEEE Conference on Computer Vision and Pattern Recognition.Columbus:IEEE,2014:580-587.
[20] REN Shaoqing,HE Kaiming,GIRSHICK R,et al.Faster R-CNN:towards real-time object detection with region proposal networks[C]//Proceedings of IEEE Transactions on Pattern Analysis and Machine Intelligence.[S.l.]:IEEE,2017:1137-1149.
[21] HE Kaiming,GKIOXARI G,DOLLÁR P,et al.Mask R-CNN[C]//Proceedings of 2017 IEEE International Conference on Computer Vision (ICCV).Venice:IEEE,2017:2980-2988.
[22] LIU Y,LU B,PENG J,ZHANG Z.Research on the use of YOLOv5 object detection algorithm in mask wearing recognition [J].World Scientific Research Journal,2020,6(11):276-284.
[23] LIU W,D.ERHAN A D,C.REED S S,et al.SSD:Single shot multibox detector[C]//Proceedings of 2016 Computer Vision-ECCV.Netherlands:Amsterdam,2016.