内河航道事故黑点识别自适应参数DBSCAN聚类算法研究  

Adaptive parameter DBSCAN clustering algorithm for inland waterway accident black spots identification

作　　者：万程鹏[1,2,3] 郭世龙 曹德胜[3] 范亮 张金奋 WAN Chengpeng;GUO Shilong;CAO Desheng;FAN Liang;ZHANG Jinfen(National Key Laboratory of Waterway Traffic Control(Wuhan University of Technology),Wuhan 430063,China;Guangdong Inland Harbor and Navigation Industry Research Co.,Ltd.,Shaoguan 511100,Guangdong,China;Intelligent Transportation Systems Research Center,Wuhan University of Technology,Wuhan 430063,China;School of Transportation and Logistics Engineering,Wuhan University of Technology,Wuhan 430063,China)

机构地区：[1]水路交通控制全国重点实验室(武汉理工大学),武汉430063 [2]广东省内河港航产业研究有限公司,广东韶关511100 [3]武汉理工大学智能交通系统研究中心,武汉430063 [4]武汉理工大学交通与物流工程学院,武汉430063

出　　处：《安全与环境学报》2024年第8期3165-3172,共8页Journal of Safety and Environment

基　　金：国家自然科学基金项目(51920105014);韶关市创新创业团队引进项目(201212176230928)。

摘　　要：内河水上交通事故时有发生,对水路运输安全、高效发展带来威胁。研究提出一种基于自适应参数的DBSCAN(Density-Based Spatial Clustering of Applications with Noise)方法,用于识别内河事故黑点水域。该方法支持对邻域半径ε和邻域中数据对象数目阈值P_(min)参数的自动选取,可提高聚类分析的精度和效率。基于2010—2019年长江干线下游散货船舶事故数据开展案例研究,对各典型事故黑点段的事故特征和事故原因进行分析,得到8个事故黑点。此外,采用Getis-Ord General G聚类识别事故黑点中的高等级事故区域,得到事故黑点及高等级事故主要分布于江心洲、桥区、港口码头区域。研究结果与实际情况基本吻合,一定程度上表明了该方法在内河水上交通事故分布特征分析上的科学性和实用性。In this study,an adaptive parameter-based DBSCAN(Density-Based Spatial Clustering of Applications with Noise)method is proposed for identifying inland river accident black spot waters.The method calculates the similarity between data points accurately,and uses the similarity calculation results for the clustering process of the DBSCAN algorithm,automatically generating the parameters of neighborhood radiusεand the threshold value of the number of data objects in the neighborhood P_(min),as well as the number of clustering clusters.When the number of generated clusters is the same three times in a row,it is considered that the clustering result tends to be stabilized,and thenεand P_(min),are regarded as the optimal clustering parameters,improving the accuracy and efficiency of the clustering analysis.Further,the Getis-Ord General G clustering method is used to analyze the indicator values in the black spots of identified accidents to obtain high-grade accident areas.Taking the accident data of downstream bulk carriers on the mainline of the Yangtze River from 2010 to 2019 as a case study,this study first conducts a preliminary statistical analysis of the data set in terms of spatial and temporal distributions,accident types,and accident grades.Subsequently,the adaptive parametric DBSCAN algorithm is used to identify the accident black spots,and eight accident black spots are obtained.Besides,three main accident types of data,namely,collision,touchdown,and self-submergence are extracted from the original dataset to identify the accident black spots of these three types of accidents.Finally,Getis-Ord General G clustering is used to analyze the accident characteristics and accident causes of each typical accident blackspot segment analysis,and it is obtained that the accident blackspots and high-level accidents are mainly distributed in the area of Jiangxinzhou,bridge area,and port terminal.

关 键 词：公共安全 交通运输安全 自适应参数DBSCAN 事故黑点 

分 类 号：X951[环境科学与工程—安全科学]