摘要
利用交通数据分析城市交通运行特征对交通规划和交通拥堵治理具有重大意义,随着车牌检测技术的发展和检测设备覆盖率的提高,车牌识别数据作为重要的交通数据来源已经在相关领域得到了广泛应用。熵作为统计学中的概念被用于交通工程中,出行熵表征了车辆运行的规律性。本文基于原始车牌识别数据提取了车辆的出行信息,在利用所得出行信息进行出行熵的模型建立时,发现由于车牌识别数据只在特定点位存在信息的特性,在计算车辆的出行熵时容易出现熵值偏大的情况,本文通过基于出行高频点位和空间距离阈值的点位聚合方法和基于出行链相似度的出行链聚合方法解决了这个问题。
Understanding urban traffic operation characteristics using traffic data is of great significance for traffic planning and traffic congestion management.With the development of license plate detection technology and the improvement of detection equipment coverage rate,license plate recognition data as an important source of traffic data has been widely used in relevant fields.Entropy,as a statistical concept,is used in traffic engineering.Travel entropy reflects the regularity of travel behavior.This paper extracts the travel information of vehicles based on the original license plate recognition data.When using the obtained travel information to establish the travel entropy model,it was found that due to the fact that license plate recognition data only has information at specific position,there is a tendency for the entropy value to be too high when calculating the travel entropy of vehicles,This paper solves this problem by using a point aggregation method based on high-frequency travel points and spatial distance thresholds,as well as a travel chain aggregation method based on travel chain similarity.
作者
王福建
洪侨波
WANG Fujian;Hong Qiaobo(IITS,College of Civil Engineering and Architecture,Zhejiang University,Hangzhou 310058,China)
出处
《综合运输》
2024年第3期114-120,共7页
China Transportation Review
基金
国家自然科学基金重点项目:基于大数据的城市交通本征获取与需求结构优化控制(52131202)。
关键词
城市交通
出行行为
熵模型
车牌识别数据
出行链
Urban transportation
Travel behavior
Entropy model
license plate recognition data
Travel chain
