摘要
提出了一种快速路入口匝道周围路径分级方法,用小波变换对交通流量数据降噪以凸显其变化特性。根据动态时间弯曲距离(DTW)比较每个交叉口进口道与入口匝道流量时间序列的相似度。应用K-means聚类方法对交叉口间路径进行分级以确定不同的控制级别。以上海市徐家汇入口匝道和周围交叉口为例,验证了本文方法的有效性。结果表明,应根据一天内不同时段交叉口到入口匝道的关键路径动态变化采取不同的控制策略。本文方法可以为不同时间段内匝道周围交叉口间关键路径识别、路径分级、区域交通信号的协调控制提供理论基础。
A route classification of intersections around expressway on-ramp is proposed.The de-noise function of Wavelet transform method is applied to mine the characteristics of traffic flow data.According to the dynamic time warping(DTW)distance,the similarity of traffic flow time series between each intersection and expressway on-ramp can be calculated.Then K-means clustering method is used to classify routes within the intersections to determine the different control levels.The Xujiahui on-ramp and adjacent intersections in Shanghai City are chosen to testify the validation of the method.The results show that key routes to the on-ramp dynamically change in different periods within a day,so different control strategies of intersections should be adopted in different periods.This method can identify key routes,achieve routes classification and provide a theoretical basis for the coordinated control of regional trafficsignals in different time periods.
作者
王文璇
朱晓东
吴兵
WANG Wen-xuan;ZHU Xiao-dong;WU Bing(Key Laboratory of Road and Traffic Engineering of Ministry of Education,Tongji University,Shanghai 201804,China;China Highway Engineering Consulting Corporation,Beijing 100089,China)
出处
《吉林大学学报（工学版）》
EI
CAS
CSCD
北大核心
2020年第3期970-979,共10页
Journal of Jilin University:Engineering and Technology Edition
基金
国家自然科学基金项目(71804127).
关键词
交通运输系统工程
路径分级
小波变换
动态时间弯曲距离
交叉口
engineering of communication and transportation system
routes classification
wavelet transform
dynamic time warping distance
intersections
