基于改进YOLOv11的光伏场区施工进度检测

doi:10.13474/j.cnki.11-2246.2025.0721

摘要/Abstract

摘要： 光伏行业作为可再生能源的重要分支,在“双碳”背景下迎来了爆发式增长。在光伏项目施工期,传统的进度管理手段已不能满足现场管理人员的需要,急需一种精确、高效的施工进度自动化管理方法。当前的目标检测算法在复杂环境下特征提取困难、检测精度低而无法适用于光伏场区的组串识别。基于此,本文提出一种光伏场区施工进度检测技术。首先,使用无人机对光伏场区进行遥感影像拍摄,并对影像数据进行自动拼接和裁剪处理;然后,使用改进YOLOv11模型对图像进行识别,识别对象包括光伏板、支架,并将识别的对象坐标信息与CAD平面布置图进行比对;最后,生成准确的场区施工进度信息用于现场管理。依托实际工程所采集的数据进行分析与验证,结果表明,该技术对于光伏板和支架的进度识别误差分别为5.2%和9.5%,与真实结果误差较小,考虑模型迭代后泛化能力的提升,可满足实际使用需求。本文使用改进的YOLOv11模型,通过引入BiFPN和SEAM模块,增强了复杂环境下的检测性能,实现了91.8%的mAP@0.5、93.4%的准确率和89.8%的召回率,相较于YOLOv11和YOLOv10模型,分别提高了1.6%、2.2%、1.2%和4.2%、3.6%、3.1%。

关键词: 进度管理, YOLOv11, BiFPN, SEAM, 光伏

Abstract: As an important branch of renewable energy,the photovoltaic industry has ushered in explosive growth under the background of “double carbon”.The traditional means of progress management during the construction period of photovoltaic projects are no longer sufficient to meet the needs of on-site managers,and there is an urgent need for an accurate and efficient construction progress automation management method.Current target detection algorithms are limited by the difficulty of feature extraction in complex environments and low detection accuracy,and can not be applied to string identification in photovoltaic fields.Based on this,this paper proposes a PV field construction progress detection technology: firstly,use a drone to take remote sensing images of the PV field,and automatically stitch and crop the image data; secondly,use the improved YOLOv11 model to identify the images,identify the objects including PV panels and racks,and compare the coordinates of the identified objects with the CAD layout plan; finally,generate the accurate construction progress information for site management.Relying on the data collected from the actual project for analysis and verification,the results show that,the progress recognition error of this technology for PV panels and brackets is5.2% and 9.5% respectively,which is smaller than the real results,and considering the improvement of the generalisation ability of the model after iteration,it can meet the requirements for practical use.The improved YOLOv11 model used in this technology enhances the detection performance in complex environments by introducing BiFPN and SEAM modules,and achieves 91.8%mAP@0.5,93.4%accuracy and 89.8%recall,which are 1.6%,2.2%,1.2%and 4.2%,3.6%,3.1%respectively compared with YOLOv11 and YOLOv10 models.

Key words: progress management, YOLOv11, BiFPN, SEAM, photovoltaic

中图分类号:

P237

魏文豪, 衷诚, 郭健辰, 张正林, 连静. 基于改进YOLOv11的光伏场区施工进度检测[J]. 测绘通报, 2025, 0(7): 132-137,163.

WEI Wenhao, ZHONG Cheng, GUO Jianchen, ZHANG Zhenglin, LIAN Jing. Improved YOLOv11-based construction progress detection technology for photovoltaic sites[J]. Bulletin of Surveying and Mapping, 2025, 0(7): 132-137,163.

参考文献

[1] 张宏霞,张衍杰,马茜,等.“双碳” 目标下新能源产业发展趋势[J].储能科学与技术,2022,11(5):1677-1678.
[2] 李行义,李海云,姚萌萌,等.基于无人机可见光影像的杆塔施工自动化进度检测技术[J].测绘通报,2021(S2):47-51.
[3] 刘莎莎,朱庆,汤圣君,等.室内三维点云与BIM集成的建筑施工进度监测方法[J].地理信息世界,2019,26(5):107-112.
[4] 罗汀,耿轶,张星.虚拟现实在机场高填方压实质量监控系统中的应用[J].工业建筑,2018,48(10):137-141.
[5] 赵顺清,孙辉.BIM+在公路工程项目管理中的应用研究[J].公路,2020,65(9):231-236.
[6] 陈钊,赵梦玲.基于BIM的水利工程施工进度实时监测系统优化设计[J].水利科技与经济,2020,26(10):102-107.
[7] 郭文斌.山地光伏电站项目施工阶段管理优化研究[D].呼和浩特:内蒙古大学,2020.
[8] 赵永强,饶元,董世鹏,等.深度学习目标检测方法综述[J].中国图象图形学报,2020,25(4):629-654.
[9] REDMON J,DIVVALA S,GIRSHICK R,et al.You only look once:unified,real-time object detection[C]//Proceedings of 2016 IEEE Conference on Computer Visionand Pattern Recognition (CVPR).LasVegas:IEEE,2016:779-788.
[10] LIU Wei,ANGUELOV D,ERHAN D,et al.SSD:single shot MultiBox detector[M]//Computer Vision - ECCV.Cham:Springer International Publishing,2016:21-37.
[11] 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,OH:IEEE,2014.
[12] REDMON J,FARHADI A.YOLO9000:better,faster,stronger[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Honolulu,HI:IEEE,2017.
[13] REDMON J,FARHADI A.YOLOv3:an incremental improvement[EB/OL].[2024-12-10].https://arxiv.org/abs/1804.02767v1.
[14] 张一帆,张梅,陈杰.基于改进YOLO v7的果蔬叶片病害检测算法[J].兰州工业学院学报,2024,31(4):48-53.
[15] 宋怀波,马宝玲,尚钰莹,等.基于YOLO v7-ECA模型的苹果幼果检测[J].农业机械学报,2023,54(6):233-242.
[16] 邓焕.基于正则化YOLO的钢表面缺陷检测方法[J].科技创新与应用,2024,14(11):168-172.
[17] GUAN Sitong,LIN Yiming,LIN Guoyu,et al.Real-time detection and counting of wheat spikes based on improved YOLOv10[J].Agronomy,2024,14(9):1936.
[18] 李姝,李思远,刘国庆.基于YOLOv8无人机航拍图像的小目标检测算法研究[J].小型微型计算机系统,2024,45(9):2165-2174.
[19] SHARMA A,KUMAR V,LONGCHAMPS L.Comparative performance of YOLOv8,YOLOv9,YOLOv10,YOLOv11 and Faster R-CNN models for detection of multiple weed species[J].Smart Agricultural Technology,2024,9:100648.
[20] 张磊,李熙尉,燕倩如,等.基于改进YOLOv5s的综采工作面人员检测算法[J].中国安全科学学报,2023,33(7):82-89.
[21] 黄玲娃,崔文成,邵虹.基于多层特征融合的行人检测方法研究[J].计算机科学,2024,51(S2):489-495.
[22] TAN Mingxing,PANG Ruoming,LE Q V.EfficientDet:scalable and efficient object detection[C]//Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).Seattle:IEEE,2020.
[23] LIN T Y,DOLLAR P,GIRSHICK R,et al.Feature pyramid networks for object detection[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Honolulu:IEEE,2017.
[24] LIU Shu,QI Lu,QIN Haifang,et al.Path aggregation network for instance segmentation[C]//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Salt Lake City:IEEE,2018.
[25] REN Yuanyuan,ZHANG Xianfeng,MA Yongjian,et al.Full convolutional neural network based on multi-scale feature fusion for the class imbalance remote sensing image classification[J].Remote Sensing,2020,12(21):3547.
[26] YU Ziping,HUANG Hongbo,CHEN Weijun,et al.YOLO-FaceV2:a scale and occlusion aware face detector[J].Pattern Recognition,2024,155:110714.
[27] 王杰.YOLO网络在焊缝检测算法中的研究与应用[D].银川:北方民族大学,2022.