Please use this identifier to cite or link to this item:
http://hdl.handle.net/10397/96801
Title: | Short-time traffic flow prediction method and device | Other Title: | 一种短时交通流量预测方法及装置 | Authors: | Shi, W Wang, R |
Issue Date: | 31-Jul-2020 | Source: | 中国专利 ZL 201710123398.1 | Abstract: | The invention is suitable for the traffic field and provides a short-time traffic flow prediction method and device. The short-time traffic flow prediction method comprises steps that a macro traffic flow model is acquired; a state vector, a state equation, an observation vector and an observation equation are determined; a data assimilation system framework for traffic flow prediction is constructed; observation data of different observation period types are classified and sampled; historical observation data are fused, and missing observation values at the present time period are completed based on the adjusted data assimilation method of set Kalman filtering; based on the data assimilation method, model parameters of a macroscopic traffic flow model are corrected and adjusted; the macroscopic traffic flow model after model parameter adjustment is utilized to predict the traffic flow in the future. The short-time traffic flow prediction method is advantaged in that the traffic flow in the future can be predicted, moreover, online adjustment is realized, and the short-time traffic flow prediction method is easy to promote. 本发明适用于交通领域,提供了一种短时交通流量预测方法及装置,所述交通流量预测方法包括:获取宏观交通流模型;确定状态向量、状态方程、观测向量和观测方程;构建用于交通流量预测的数据同化系统框架;将不同观测时段类型的观测数据进行分类采样;融合历史观测数据,基于调整的集合卡尔曼滤波的数据同化方法,补齐当前时刻路段缺失的观测值;基于所述数据同化方法,对所述宏观交通流模型的模型参数进行修正调整;利用调整模型参数后的所述宏观交通流模型,对未来时刻的交通流量进行预测;本发明能够对未来时刻的交通流量进行预测,同时实现了在线调整,易于推广。 |
Publisher: | 中华人民共和国国家知识产权局 | Rights: | Assignee: 香港理工大学深圳研究院 |
Appears in Collections: | Patent |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
ZL201710123398.1.PDF | 739.17 kB | Adobe PDF | View/Open |
Page views
14
Citations as of May 5, 2024
Downloads
5
Citations as of May 5, 2024
Google ScholarTM
Check
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.