The invention discloses an urban road network dynamic OD estimation method based on transfer learning. The research object of the method is the dynamic traffic demand in an urban road network, the basic data of the method comprises social vehicle license plate recognition data and taxi GPS data, and supervised dynamic OD estimation is completed based on a transfer learning framework. Considering that the travel of a taxi in a passenger carrying state is similar to the travel of a social vehicle, effective isomorphic travel travel data is extracted from license plate data and GPS data, and travel origin and destination identification knowledge in taxi data is applied to the license plate data of the social vehicle by applying transfer learning. And determining starting and ending points of social vehicle travel so as to complete dynamic OD estimation.
本发明公开了一种基于迁移学习的城市路网动态OD估计方法。本发明的研究对象是城市路网中的动态交通需求,方法的基础数据包括社会车辆车牌识别数据以及出租车GPS数据,基于迁移学习的框架完成有监督的动态OD估计。考虑到载客状态下的出租车与社会车辆的行程具有相似性,分别从号牌数据和GPS数据中提取有效的同构出行行程数据,应用迁移学习将出租车数据中的出行起讫点识别知识应用于社会车辆号牌数据,确定社会车辆出行的起终点从而完成动态OD估计。
Urban road network dynamic OD estimation method based on transfer learning
一种基于迁移学习的城市路网动态OD估计方法
2023-07-21
Patent
Electronic Resource
Chinese
IPC: | G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS |
Urban road network unknown flow estimation method and system based on transfer learning model
European Patent Office | 2021
|European Patent Office | 2023
|Full-road-network road section travel time estimation method based on urban road checkpoint data
European Patent Office | 2020
|Urban road network dynamic OD estimation method based on mobile crowdsourcing and AVI data
European Patent Office | 2022
|Novel urban road network dynamic subarea division method
European Patent Office | 2023
|