摘要
针对高速公路站级联网收费流水数据,从流量、车辆和交通状态等3个视角建立数据分析框架。以广东省某收费站3个月的通行流水数据为例,提取流量的宏观分布、车辆使用频度和交通状态等特征,并采用时间序列聚类算法进一步挖掘流量的时间分布规律。结果表明,基于流量、车辆和交通状态等3个视角建立的分析框架能有效提取原始数据中的公路交通运行信息,适用于自由流收费制式下的流水数据挖掘。
Deep analysis and research of massive data from highway electro-mechanical system will play an indispensible role in developing smart highways.This paper proposes an analysis framework for traffic volume,vehicle types and traffic situation.The data from a toll station in Guangdong province for 3 months are processed for illustration.The macro-distribution of traffic volumes,the occurring frequency of different types of vehicles and the features of traffic situation are extracted and a time series clustering is conducted to find the daily-distribution law of the traffic volume.The analysis framework will apply to future toll collection mode of free flow as well.
作者
崔毓伟
李隆杰
CUI Yuwei;LI Longjie(COSCO SHIPPING Technology Co.,Ltd.,Shanghai 200135,China)
出处
《上海船舶运输科学研究所学报》
2019年第4期55-60,共6页
Journal of Shanghai Ship and Shipping Research Institute
关键词
智慧高速
联网收费数据
特征提取
时间序列聚类
smart highway
highway network toll data
feature extraction
time series clustering
