|本期目录/Table of Contents|

[1]李文涛,韩豫,杨林,等.施工空间冲突事故险兆特征智能检测方法*[J].中国安全生产科学技术,2023,19(11):157-165.[doi:10.11731/j.issn.1673-193x.2023.11.022]
　LI Wentao,HAN Yu,YANG Lin,et al.Intelligent detection method for near-miss characteristics of construction space conflict accidents[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2023,19(11):157-165.[doi:10.11731/j.issn.1673-193x.2023.11.022]
点击复制

施工空间冲突事故险兆特征智能检测方法*
分享到：

《中国安全生产科学技术》[ISSN:1673-193X/CN:11-5335/TB]

卷:
19
期数:
2023年11期
页码:
157-165
栏目:
职业安全卫生管理与技术
出版日期:
2023-11-30

文章信息/Info

Title:
Intelligent detection method for near-miss characteristics of construction space conflict accidents
文章编号:
1673-193X(2023)-11-0157-09
作者:
李文涛; 韩豫; 杨林; 李康
（1.江苏大学 土木工程与力学学院，江苏 镇江 212013；
2.江苏大学 应急管理学院，江苏 镇江 212013）
Author(s):
LI Wentao;  HAN Yu;  YANG Lin;  LI Kang
(1.Faculty of Civil Engineering and Mechanics,Jiangsu University,Zhenjiang Jiangsu 212013,China;
2.School of Emergency Management,Jiangsu University,Zhenjiang Jiangsu 212013,China)
关键词:
施工安全; 空间冲突; 深度学习; 空间测距; 事故险兆
Keywords:
construction safety;  space conflict;  deep learning;  space distance measurement;  accident near-miss
分类号:
X947
DOI:
10.11731/j.issn.1673-193x.2023.11.022
文献标志码:
A
摘要:
为了更及时、准确地发现施工场景中潜在空间事故要素，减少相关事故发生。提出1种结合视觉检测与透视变换下的图像距离估计技术的施工空间冲突事故险兆特征检测方法，该方法以YOLOv7为基础模型，将主干网络替换为轻量级MobileOne，同时在特征融合网络与检测头中分别加入SPD-Conv卷积构建块及SimAM无参数注意力机制。实现目标检测、运动状态判别、距离测算3大功能模块的整合，感知空间冲突事故险兆特征进而及时阻断事故演进。研究结果表明：此方法在险兆目标检测任务中相较原始网络在mAP精度小幅降低2.79个百分点的情况下检测速度提升了50%；其次，运动属性险兆特征平均识别准确率达91%以上，测距精度保持在92%~96%。研究结果可为实现施工场景空间冲突险兆特征的有效检测提供参考。
Abstract:
In order to discover the potential spatial accident elements in the construction scene more timely and accurately,and reduce the occurrence of related accidents,a detection method for the near-miss characteristics of construction space conflict accidents combining with the image distance estimation technology based on visual detection and perspective transformation was proposed.Taking YOLOv7 as the basic model,the backbone network was replaced with lightweight MobileOne.At the same time,the SPD-Conv convolution building blocks and SimAM parameter-free attention mechanism were added to the feature fusion network and the detection head,respectively.The integration of three functional modules of target detection,motion state discrimination and distance measurement was realized,and the near-miss characteristics of space conflict accidents were perceived,so as to block the accident evolution in time.The results show that compared with the original network,the detection speed of this method is increased by 50% when the mAP accuracy is slightly reduced by 2.79% in the task of near-miss target detection.Secondly,the average recognition accuracy of motion attribute near-miss characteristics reaches above 91%,and the distance measurement accuracy is maintained in the range of 92%~96%.The research results provide a reference for the effective detection of spatial conflict near-miss characteristics in construction scenes.

参考文献/References:

［1］王珊珊,王伟.AP1000重要模块运输吊装安全措施研究［J］.中国安全生产科学技术,2016,12(增刊1):97-102. WANG Shanshan,WANG Wei.Study on safety measures for transporting and lifting of AP1000 important modules［J］.Journal of Safety Science and Technology,2016,12(Supplement 1):97-102.
［2］陈丽娟.基于BIM的地铁施工空间安全管理研究［D］.武汉：华中科技大学,2012.
［3］马辉,张文静,董美红.装配式建筑吊装施工空间冲突分析与多目标优化［J］.中国安全科学学报,2020,30(2):28-34. MA Hui,ZHANG Wenjing,DONG Meihong.Spatial conflict analysis and multi-objective optimization for prefabricated building hoisting construction［J］.China Safety Science Journal,2020,30(2):28-34.
［4］SUGIMOTO Y,SEKI H,SAMO T,et al.4D CAD-based evaluation system for crane deployment plans in construction of nuclear power plants［J］.Automation in Construction,2016,71:87-98.
［5］马辉,张红滨,王艳娜.基于BIM的装配式建筑并行施工作业空间冲突识别［J］.中国安全生产科学技术,2020,16(2):97-103. MA Hui,ZHANG Hongbin,WANG Yanna,Conflict recognition on working space in concurrent construction of prefabricated building based on BIM［J］.Journal of Safety Science and Technology,2020,16(2):97-103.
［6］ZHANG C,HAMMAD A,RODRIGUEZ S.Crane pose estimation using UWB real-time location system［J］.Journal of Computing in Civil Engineering,2012,26(5):625-637.
［7］陈浩,李起伟,王子龙.基于改进TDOA在煤矿井下超宽带定位算法的研究［J］.电子测量技术,2021,44(6):96-102. CHEN Hao,LI Qiwei,WANG Zilong,Based on improved TDOA in underground mine research on UWB location algorithm［J］.Electronic Measurement Technology,2021,44(6):96-102.
［8］KIM K,KIM M.RFID-based location-sensing system for safety management［J］.Personal and Ubiquitous Computing,2012,16:235-243.
［9］杨荣坚,王芳,秦浩.基于双目图像的行人检测与定位系统研究［J］.计算机应用研究,2018,35(5):1591-1595,1600. YANG Rongjian,WANG Fang,QIN Hao.Research of pedestrian detection and location system based on stereo images［J］.Application Research of Computers,2018,35(5):1591-1595,1600.
［10］XIANG X,LYU N,GUO X,et al.Engineering vehicles detection based on modified faster R-CNN for power grid surveillance［J］.Sensors,2018,18(7):2258.
［11］石怀涛,李刚,范丽婷,等.吊车危险区域侵入行人机器视觉定位算法［J］.控制工程,2021,28(4):759-765. SHI Huaitao,LI Gang,FAN Liting,et al.Machine visual positioning algorithm for pedestrians entering dangerous areas of cranes［J］.Control Engineering of China,2021,28(4):759-765.
［12］岳晨曦.基于事故致因理论的地铁施工险兆事件安全管理研究［D］.西安：西安科技大学,2020.
［13］WANG C Y，BOCHKOVSKIY A,LIAO H Y M.YOLOv7:trainable bag-of-freebies sets new state-of-the-art for real-time object detectors［C］//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Vancouver：Computer Vision Foundation，2023.
［14］VASU P K A,GABRIEL J,ZHU J,et al.MobileOne:An Improved One Millisecond Mobile Backbone［C］//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Vancouver：Computer Vision Foundation，2023.
［15］HURTIK P,MOLEK V,HULA J,et al.Poly-YOLO:higher speed,more precise detection and instance segmentation for YOLOv3［J］.Neural Computing and Applications,2022,34(10):8275-8290.
［16］王坚,彭晖,文俊,等.基于多视角特征点匹配的室外目标定位［J］.华中科技大学学报(自然科学版),2015,43(增刊1):241-244,249. WANG Jian,PENG Hui,WEN Jun,et al.Outdoor object localization based on multiple views of feature points matching［J］.Journal of Huazhong University of Science and Technology (Natural Science Edition),2015,43(Supplement 1):241-244,249.
［17］刘亚坤,李永强,刘会云，等.基于改进RANSAC算法的复杂建筑物屋顶点云分割［J］.地球信息科学学报,2021,23(8):1497-1507. LIU Yakun,LI Yongqiang,LIU Huiyun,et al.An improved RANSAC algorithm for point cloud segmentation of complex building roofs［J］.Journal of Geo-Information Science，2021,23(8):1497-1507.

相似文献/References:

[1]宋四新 ,宋先勇.大型排架施工安全监控机制研究[J].中国安全生产科学技术,2012,8(6):101.
　SONG Si xin,SONG Xian yong.Research on safety monitoring mechanism of largescale bent frame construction[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2012,8(11):101.
[2]鲍学英,王起才,宫文昌.基于模糊层次分析法的地铁工程施工安全评价[J].中国安全生产科学技术,2013,9(1):136.
　BAO Xue ying,WANG Qi cai.Study on safety assessments of metro constructions using fuzzy analytic hierarchy process[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2013,9(11):136.
[3]黄丽霞,张东曾.悬索桥钢箱梁的吊装施工安全技术研究[J].中国安全生产科学技术,2010,6(3):228.
　HUANG Li-xia,ZHANG Dong-zeng.Study on the safe hoisting techniques for steel box girder of suspension bridge[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2010,6(11):228.
[4]袁春燕,杨剑耕,卢俊龙.模糊测度理论在施工现场安全目标决策中的应用[J].中国安全生产科学技术,2014,10(3):101.[doi:10.11731/j.issn.1673-193x.2014.03.017]
　YUAN Chun yan,YANG Jian geng,LU Jun long.Application of fuzzy measure theory in safety objective decision of construction site[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2014,10(11):101.[doi:10.11731/j.issn.1673-193x.2014.03.017]
[5]杨高升,李松晨,姜双双.建筑施工企业安全激励机制优化研究[J].中国安全生产科学技术,2014,10(6):133.[doi:10.11731/j.issn.1673-193x.2014.06.021]
　YANG Gao sheng,LI Song chen,JIANG Shuang shuang.Research on optimization of safety incentive mechanism in construction enterprises[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2014,10(11):133.[doi:10.11731/j.issn.1673-193x.2014.06.021]
[6]郭红领,张伟胜,刘文平.集成设计-施工安全（DFCS）的方法探索[J].中国安全生产科学技术,2015,11(2):5.[doi:10.11731/j.issn.1673-193x.2015.02.001]
　GUO Hong-ling,ZHANG Wei-sheng,LIU Wen-ping.Exploration on integration method of design for construction safety(DFCS)[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2015,11(11):5.[doi:10.11731/j.issn.1673-193x.2015.02.001]
[7]韩豫,张泾杰,梅强,等.建筑工人安全行为习惯的塑造策略与方法[J].中国安全生产科学技术,2015,11(9):177.[doi:10.11731/j.issn.1673-193x.2015.09.028]
　HAN Yu,ZHANG Jing-jie,MEI Qiang,et al.Shaping strategies and methods of safety behavior habits for construction workers[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2015,11(11):177.[doi:10.11731/j.issn.1673-193x.2015.09.028]
[8]郑霞忠,陈云,陈述.溧阳电站整流锥支撑系统施工安全影响因素识别研究[J].中国安全生产科学技术,2016,12(3):137.[doi:10.11731/j.issn.1673-193x.2016.03.025]
　ZHENG Xiazhong,CHEN Yun,CHEN Shu.Recognition of influence factors on construction safety for support system of air flow distributing wimble in Liyang power station[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2016,12(11):137.[doi:10.11731/j.issn.1673-193x.2016.03.025]
[9]韩豫,梅强,周丹,等.群体封闭性视角下的建筑工人不安全行为传播特[J].中国安全生产科学技术,2016,12(3):187.[doi:10.11731/j.issn.1673-193x.2016.03.034]
　HAN Yu,MEI Qiang,ZHOU Dan,et al.Propagation characteristics of unsafe behaviors for construction workers from the perspective of group closeness[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2016,12(11):187.[doi:10.11731/j.issn.1673-193x.2016.03.034]
[10]韩豫,张泾杰,孙昊,等.基于图像识别的建筑工人智能安全检查系统设计与实现[J].中国安全生产科学技术,2016,12(10):142.[doi:10.11731/j.issn.1673-193x.2016.10.024]
　HAN Yu,ZHANG Jingjie,SUN Hao,et al.Design and implementation of intelligent safety inspection system for construction workers based on image recognition[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2016,12(11):142.[doi:10.11731/j.issn.1673-193x.2016.10.024]

备注/Memo

备注/Memo:
收稿日期： 2023-05-20
* 基金项目： 国家自然科学基金项目（72071097）；教育部人文社会科学研究规划基金项目（20YJAZH034）；2022年江苏省研究生实践创新计划项目（JSCX22_1872）；江苏大学应急管理学院专项科研项目（KY-B-10）
作者简介： 李文涛，硕士研究生，主要研究方向为智慧建造。
通信作者： 韩豫，博士，教授，主要研究方向为智慧建造与施工安全。
更新日期/Last Update: 2023-12-06