计算机应用研究2023，Vol.40Issue(12)：3578-3583,6.DOI:10.19734/j.issn.1001-3695.2023.04.0149
基于改进密度峰值聚类的路网划分方法
Road network classification method based on improved density peak clustering
摘要
Abstract
The reasonable classification of urban road network is of great significance for the optimization of regional traffic control and the implementation of coordination strategies.In order to improve road traffic efficiency,this paper proposed an ur-ban road network classification method based on density peak clustering algorithm.Firstly,it comprehensively considered the influence of static and dynamic factors at intersections to construct a correlation degree model for adjacent intersections,and provided a quantitative description for reasonably quantifying the correlation degree between intersections.Secondly,it proposed an improved density peak clustering algorithm that combined the correlation between adjacent intersections to partition the road network area.To address the problem that the local density in the density peak clustering algorithm varies greatly on different size data sets,it introduced the idea of KNN to re-describe the local density,and secondly,in order to avoid the subjectivity of the manual selection of the algorithm clustering center which could lead to error problem,using the elbow rule to realize the au-tomatic selection of the clustering center.The experimental results show that compared with the improved Newman algorithm and Ncut algorithm,the improved proposed algorithm can reduce 12.5％and 22.8％respectively in optimizing the flat homo-geneity of sub-regions,improve the effect of control sub-region division and make the region division effect more reasonable.关键词
区域划分/交叉口关联度/密度峰值聚类算法/KNN/肘部法则Key words
regional division/intersection correlation/density peak clustering/KNN/elbow law分类
信息技术与安全科学引用本文复制引用
杨迪,徐文瑜,王鹏..基于改进密度峰值聚类的路网划分方法[J].计算机应用研究,2023,40(12):3578-3583,6.基金项目
吉林省教育厅科学研究项目(JJKH20230848KJ) （JJKH20230848KJ）
吉林省科技创新平台建设项目(YDZJ202302CXJD027) （YDZJ202302CXJD027）
