Enterprise safety risk management model of machinery manufacturing industry based on improved YOLOv5

doi:10.16265/j.cnki.issn1003-3033.2025.03.0471

Abstract

Abstract:

In order to improve the efficiency and accuracy of safety risk management in machinery manufacturing enterprises, the Bayesian network and machine vision technology were combined. Based on improved YOLOv5, Intersection over Union(IoU) values of safety hazard events occurring at the operation site were calculated. By leveraging the audit risk assessment in conjunction with AHP to derive the danger weights, the prior probabilities of the root nodes of Bayesian network were determined. Bayesian network model and design management system were established to realize closed-loop control. A safety risk management model of machinery manufacturing enterprises was constructed and verified by examples. The results show that the model has a more accurate identification and evaluation ability, and can find some potential safety hazards, so as to optimize the current management process. At the same time, the model also successfully realizes the effective combination of qualitative and quantitative analysis, integrates the expert experience and data quantification results, and confirms each other, so that the risk assessment results have a certain improvement in scientificity and reliability, which can provide a practical new idea for safety risk management.

Key words: machinery manufacturing industry, safety risk management, Bayesian network, machine vision, improved YOLOv5, analytic hierarchy process (AHP), prior probability

CLC Number:

X941

ZHANG Hao, HAIMUDULA Aierken. Enterprise safety risk management model of machinery manufacturing industry based on improved YOLOv5[J]. China Safety Science Journal, 2025, 35(3): 52-59.

Figures/Tables 16

Fig.1

Fig.2

Fig.3

Table 1

Table 2

Table 3

Table 4

Table 5

Table 6

Table 7

Table 8

Table 9

Table 10

Table 11

Fig.4

Fig.5

References 11

[1]	AMIN T, KHAN F, IMTIAZ S. Dynamic availability assessment of safety critical systems using a dynamic bayesian network[J]. Reliability Engineering and System Safety, 2018,178: 108-117.
[2]	高萦莹. 基于计算机视觉数据驱动的新型施工安全风险评估研究[D]. 大连: 东北财经大学, 2023.
	GAO Yingying. Research on newvonstruction safety risk assessment driven by computer vision data[D]. Dalian: Dongbei University of Finance and Economics, 2023.
[3]	杜敏维, 于德湖, 邵志国. 基于T-S模糊故障树和贝叶斯网络的建筑施工风险评估模型建立及应用[J]. 青岛:青岛理工大学学报, 2023, 44(1):21-29.
	DU Minwei, YU Dehu, SHAO Zhiguo. Establishment and application of construction risk assessment model based on T-S fuzzy fault tree and bayesian network[J]. Journal of Qingdao University of Technology, 2023, 44(1):21-29.
[4]	吕艳, 古丽赛乃穆·阿布都热合曼,高程. 网红景点安全风险管理初探[J]. 灾害学, 2024, 39(3):132-138.
	LYU Yan, ABUDUREHEMAN Gulisainaimu, GAO Cheng. A preliminary study on safety risk management of Internet-famous tourist attractions[J]. Journal of Catastrophology, 2024, 39 (3): 132-138.
[5]	王金江, 关鹏婷, 陈卓, 等. 基于深度学习的检修作业过程风险智能预警[J]. 中国安全科学学报, 2023, 33(10): 16-22. doi: 10.16265/j.cnki.issn1003-3033.2023.10.1744
	WANG Jinjiang, GUAN Pengting, CHEN Zhuo, et al. Intelligent early warning of risks in the maintenance operation process based on deep learning[J]. China Safety Science Journal, 2023, 33 (10): 16-22. doi: 10.16265/j.cnki.issn1003-3033.2023.10.1744
[6]	唐哲, 刘艺超, 方汀, 等. 基于YOLOv5和特征提取机制的作业穿戴装备检测方法研究[J]. 科学技术创新, 2023(8): 48-51.
	TANG Zhe, LIU Yichao, FANG Ting, et al. Research on the detection method of work wearing equipment based on YOLOv5 and feature extraction mechanism[J]. Scientific and Technological Innovation, 2023 (8): 48-51.
[7]	魏斌庆. 基于文本和图像信息的建筑施工现场险兆事件管理研究[D]. 徐州: 中国矿业大学, 2023.
	WEI Binqing. Research on the management of near-miss events in construction sites based on text and image information[D]. Xuzhou: China University of Mining and Technology, 2023.
[8]	魏峰, 周建平, 谭翔, 等. 针对低空微小型无人机的轻量型YOLOv5检测算法[J]. 光电子·激光, 2024, 35(6): 641-649.
	WEI Feng, ZHOU Jianping, TAN Xiang, et al. Lightweight YOLOv5 detection algorithm for low-altitude micro and small unmanned aerial vehicles[J]. Journal of Optoelectronics·Laser, 2024, 35 (6): 641-649.
[9]	崔雅欣. 基于故障树和贝叶斯网络的建筑施工高处坠落事故致因分析[D]. 西安: 西安理工大学, 2023.
	CUI Yaxin. Cause analysis of construction falls accidents based on fault Tree and bayesian network[D]. Xi'an: Xi'an University of Technology, 2023.
[10]	马港, 徐晓楠, 郭小芳, 等. 基于贝叶斯网络的电镀企业火灾事故情景推演[J]. 中国安全科学学报, 2023, 33(2): 202-208. doi: 10.16265/j.cnki.issn1003-3033.2023.02.0420
	MA Gang, XU Xiaonan, GUO Xiaofang, et al. Scenario deduction of fire accidents in electroplating enterprises based on Bayesian network[J]. China Safety Science Journal, 2023, 33 (2): 202-208. doi: 10.16265/j.cnki.issn1003-3033.2023.02.0420
[11]	李博宇. 铁路技术站生产作业智能化安全管理及评价技术研究[D]. 北京: 中国铁道科学研究院, 2023.
	LI Boyu. Research on intelligent safety management and evaluation technology for production operations at railway technical stations[D]. Beijing: China Academy of Railway Sciences, 2023.

模型	F₁	mAP@0.5
YOLOv5-/NL(原模型)	0.87	0.919
YOLOv5-/NS	0.96	0.967
YOLOv5-/SS	0.85	0.917
YOLOv5-/SL	0.97	0.966
YOLOv5-SSL	0.96	0.964
YOLOv5-SNL	0.97	0.972
YOLOv5-SNS	0.97	0.980
YOLOv5-SSS	0.97	0.983

标度	具体含义
1	2指标相比同等重要
3	2指标相比,前者比后者稍微重要
5	2指标相比,前者比后者明显重要
7	2指标相比,前者比后者强烈重要
9	2指标相比,前者比后者极端重要
2、4、6、8	2指标相比,前者比后者重要的程度在上述之间
1/3、1/5、1/7、1/9	2指标相比,后者比前者重要的上述相应程度
1/2、1/4、1/6、1/8	2指标相比,后者比前者重要的程度在上述之间

隐患事件	环境复杂程度	事故后果严重性	历史事故发生次数	人员密集度	应急响应有效性
环境复杂程度	1	1/7	1/5	1/4	1/6
事故后果严重性	7	1	5	6	3
历史事故发生次数	5	1/5	1	3	2
人员密集程度	4	1/6	1/3	1	1/4
应急响应的有效性	6	1/3	1/2	4	1

隐患事件	环境复杂度	事故后果严重性	历史事故发生次数	人员密集度	应急响应有效性	O	Aw
环境复杂度	0.043	0.078	0.028	0.018	0.026	0.038 6	0.20
事故后果严重性	0.304	0.543	0.711	0.421	0.467	0.489 5	2.83
历史事故发生次数	0.217	0.109	0.142	0.211	0.312	0.198 1	1.12
人员密集度	0.174	0.091	0.047	0.070	0.039	0.084 2	0.43
应急响应有效性	0.261	0.181	0.071	0.281	0.156	0.189 9	1.02

检验结果	准则层
λ_max	5.432 089 644
CI	0.108 0
RI	1.12
CR	0.096