|本期目录/Table of Contents|

[1]贺姝逸,罗杰,刘顶立,等.基于PCA与K-means聚类的多维火灾风险空间分异分析*[J].中国安全生产科学技术,2025,21(11):82-89.[doi:10.11731/j.issn.1673-193x.2025.11.010]
　HE Shuyi,LUO Jie,LIU Dingli,et al.Multidimensional fire risk spatial differentiation analysis based on PCA and K-means clustering[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2025,21(11):82-89.[doi:10.11731/j.issn.1673-193x.2025.11.010]
点击复制

基于PCA与K-means聚类的多维火灾风险空间分异分析*
分享到：

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

卷:
21
期数:
2025年11期
页码:
82-89
栏目:
职业安全卫生管理与技术
出版日期:
2025-11-30

文章信息/Info

Title:
Multidimensional fire risk spatial differentiation analysis based on PCA and K-means clustering
文章编号:
1673-193X(2025)-11-0082-08
作者:
贺姝逸; 罗杰; 刘顶立; 刘伟军; 卜蓉伟; 颜龙
（1.长沙理工大学 交通学院，湖南 长沙 410114；
2.邵阳市消防救援支队，湖南 邵阳 422004；
3.中南大学 土木工程学院，湖南 长沙 410075)
Author(s):
HE Shuyi;  LUO Jie;  LIU Dingli;  LIU Weijun;  BU Rongwei;  YAN Long
(1.School of Transportation,Changsha University of Science & Technology,Changsha Hunan 410114,China;
2.Shaoyang Fire and Rescue Division,Shaoyang Hunan 422004,China;
3.School of Civil Engineering,Central South University,Changsha Hunan 410075,China)
关键词:
区域火灾风险; PCA算法; K-means聚类算法; 空间分异
Keywords:
regional fire risk;  PCA algorithm;  K-means clustering algorithm;  spatial differentiation
分类号:
X932
DOI:
10.11731/j.issn.1673-193x.2025.11.010
文献标志码:
A
摘要:
为揭示区域火灾风险的空间分异规律，提出构建涵盖社会经济、人口、地理及火灾特征的多维指标体系。基于邵阳市2021—2024年火灾数据，采用PCA提取核心主成分，结合K-means聚类对主成分得分进行空间分区。探究各类区域侧重的风险驱动因子，并结合场景信息及火灾风险空间热力图对模型的有效性进行验证。研究结果表明：由于各区域功能及地理位置不同，火灾风险特征存在明显差异，模型划分出4类特征鲜明的风险区域：均衡复合型、低风险约束型、经济热点型、自然隐患型，与热力图展示的“东高-中稳-西险”空间梯度高度吻合。研究结果证实了PCA与K-means聚类的多维区域火灾风险模型的有效性，可为区域差异化消防治理提供参考。
Abstract:
In order to reveal the spatial differentiation patterns of regional fire risk,a multidimensional indicator system that covers socioeconomic,demographic,geographic,and fire-related characteristics is constructed.Based on fire incident data for Shaoyang City from 2021 to 2024,principal component analysis (PCA) is used to extract core principal components,and K-means clustering is combined with the principal component scores to perform spatial partitioning.The dominant risk-driving factors emphasized in each type of region are investigated,and the effectiveness of the model is validated using scenario information and spatial heat maps of fire risk.The results show that,due to differences in regional functions and geographical locations,the characteristics of fire risk differ markedly,and the model delineates four distinct types of risk regions-balanced composite type,low-risk constrained type,economic hotspot type,and natural-hazard-prone type-which are highly consistent with the “east-high-central-stable-west-risky” spatial gradient displayed by the heat maps.The findings confirm the effectiveness of the multidimensional regional fire risk model integrating PCA and K-means clustering and can provide a reference for differentiated fire protection governance across regions.

参考文献/References:

［1］王倩倩,孟繁宇,曾益萍,等.基于DBSCAN聚类的城市区域火灾风险计算方法——以深圳市盐田区为例［J］.中国安全生产科学技术,2021,17(2):177-182. WANG Qianqian,MENG Fanyu,ZENG Yiping,et al.Calculation method of fire risk for urban areas based on DBSCAN clustering:a case study of Yantian District,Shenzhen ［J］.Journal of Safety Science and Technology,2021,17(2):177-182.
［2］国家消防救援局.【央视新闻】全年接报90.8万起 消防部门盘点2024全国火灾五个特点［EB/OL］.( 2025-01-24)［ 2025-05-31］.https://www.119.gov.cn/site1/qmxfxw/mtbd/wzbd/2025/48022.shtml.
［3］LEE H J,PARK S W,ROH S,et al.Deriving major fire risk evaluation items utilizing spatial information convergence technology in dense areas of small obsolete buildings ［J］.Sustainability,2021,13(22):12593.
［4］刘顶立,徐志胜,王飞跃.基于后果概率估计的地铁集群火灾风险评估［J］.铁道科学与工程学报,2021,18(3):813-820. LIU Dingli,XU Zhisheng,WANG Feiyue.Fire risk evaluation of subway group based on consequence probability estimation ［J］.Journal of Railway Science and Engineering,2021,18(3):813-820.
［5］QI C M,ZOU Q F,CAO Y,et al.Hazardous chemical laboratory fire risk assessment based on ANP and 3D risk matrix ［J］.Fire,2024,7(8):287.
［6］ZHANG R J,TIAN D,WANG H W,et al.Risk assessment of compound dynamic disaster based on AHP-EWM ［J］.Applied Sciences,2023,13(18):10137.
［7］姜洪,曾胜,周添,等.基于改进AHP和熵权法的特长公路隧道消防救援能力评价［J］.工业安全与环保,2025,51(2):28-33. JIANG Hong,ZENG Sheng,ZHOU Tian,et al.Evaluation of fire service capability of extra-long highway tunnels based on improved AHP and entropy weight method ［J］.Industrial Safety and Environmental Protection,2025,51(2):28-33.
［8］BENTO V A,LIMA D C A,SANTOS L C,et al.The future of extreme meteorological fire danger under climate change scenarios for Iberia ［J］.Weather and Climate Extremes,2023,42:100623.
［9］ZHENG Q,ZHOU A,SHEN S L.Mapping bushfire risk based on scale division and factor analysis:a case study from Victoria,Australia ［J］.International Journal of Disaster Risk Reduction,2025,118:105222.
［10］陈志芬,黄靖玲,李亚.适应城市消防规划需求的火灾风险评估研究［J］.中国安全生产科学技术,2019,15(5):185-191. CHEN Zhifen,HUANG Jingling,LI Ya.Study on fire risk assessment adapting to requirements of urban fire protection planning ［J］.Journal of Safety Science and Technology,2019,15(5):185-191.
［11］YUN B S,XU X H,ZENG J,et al.An improved unmanned aerial vehicle forest fire detection model based on YOLOv8 ［J］.Fire,2025,8(4):138.
［12］WANG C,LIU H Z,XU Y Q,et al.A forest fire prediction framework based on multiple machine learning models ［J］.Forests,2025,16(2):329.
［13］DUO M,HU J,FANG Z T,et al.Fire risk in the context of social development and government control:evidence from 10 years of multivariate statistics in China ［J］.Fire Safety Journal,2025,153:104383.
［14］ARREERAS S,PHONSITTHANGKUN S,ARREERAS T,et al.Spatial analysis on the service coverage of emergency facilities for fire disaster risk in an urban area using a web scraping method:a case study of Chiang Rai City,Thailand ［J］.Urban Science,2024,8(3):140.
［15］中国工程建设标准化协会.区域火灾风险评估技术规程：T/CECS 887—2021［S］.北京:中国建筑工业出版社,2021.
［16］LIU D L,XU Z S,FAN C G,et al.Development of fire risk visualization tool based on heat map ［J］.Journal of Loss Prevention in the Process Industries,2021,71:104505.

相似文献/References:

[1]李垣志,牛国庆,刘慧玲.改进的GA-BP神经网络在矿井突水水源判别中的应用[J].中国安全生产科学技术,2016,12(7):77.[doi:10.11731/j.issn.1673-193x.2016.07.014]
　LI Yuanzhi,NIU Guoqing,LIU Huiling.Application of improved GA-BP neural network on identification of water inrush source in mine[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2016,12(11):77.[doi:10.11731/j.issn.1673-193x.2016.07.014]
[2]王倩倩,孟繁宇,曾益萍,等.基于DBSCAN聚类的城市区域火灾风险计算方法——以深圳市盐田区为例*[J].中国安全生产科学技术,2021,17(2):177.[doi:10.11731/j.issn.1673-193x.2021.02.028]
　WANG Qianqian,MENG Fanyu,ZENG Yiping,et al.Calculation method of fire risk for urban areas based on DBSCAN clustering:A case study of Yantian District,Shenzhen[J].JOURNAL OF SAFETY SCIENCE AND TECHNOLOGY,2021,17(11):177.[doi:10.11731/j.issn.1673-193x.2021.02.028]

备注/Memo

备注/Memo:
收稿日期： 2025-08-14
* 基金项目： 国家自然科学基金青年科学基金项目（52204202，52504201）
作者简介： 贺姝逸，硕士研究生，主要研究方向为火灾风险防控。
通信作者： 刘顶立，博士，副教授，主要研究方向为火灾风险防控及消防救援。
更新日期/Last Update: 2025-12-03