Unlock instant, AI-driven research and patent intelligence for your innovation.

Method and system for traffic forecasting of highway sites based on spatiotemporal attention mechanism

A technology of flow forecasting and attention, applied in forecasting, computer-aided design, instruments, etc., can solve problems such as inaccurate forecasting and failure to capture the dynamic spatial correlation of toll stations, and achieve the effect of improving forecasting accuracy

Active Publication Date: 2022-08-02
CENT SOUTH UNIV
View PDF9 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The present invention provides a method and system for traffic forecasting of highway website sites based on a spatio-temporal attention mechanism, which is used to solve the technical problem of inaccurate predictions caused by the inability of existing traffic forecasting methods to capture the dynamic spatial correlation of toll sites

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Method and system for traffic forecasting of highway sites based on spatiotemporal attention mechanism
  • Method and system for traffic forecasting of highway sites based on spatiotemporal attention mechanism
  • Method and system for traffic forecasting of highway sites based on spatiotemporal attention mechanism

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0050] This implementation discloses a method for predicting traffic on a highway site based on a spatiotemporal attention mechanism, which includes the following steps:

[0051] Build a training sample set of the target site, the training sample set includes the historical outbound traffic sequence corresponding to the predicted time of the target site, the historical time feature vector corresponding to the historical outbound traffic sequence, the time feature vector of the predicted time of the target site and the corresponding The feature vector of inbound traffic at the time associated with the key source site;

[0052] Build a site traffic prediction model based on the spatiotemporal attention mechanism, and use the training data in the training sample set to train the site traffic prediction model, so that it learns the relationship between the traffic data at the target site's predicted moment and its historical traffic data in the past period of time. and the dynamic...

Embodiment 2

[0056] The second embodiment is a preferred embodiment of the first embodiment. The difference from the first embodiment is that the specific steps of the method for predicting the traffic of highway sites based on the spatiotemporal attention mechanism are expanded:

[0057] In this embodiment, a method for predicting the traffic of expressway stations based on the spatiotemporal attention mechanism is disclosed, which can effectively capture the time dependence and dynamic space of the traffic of expressway toll stations through the temporal attention and spatial attention mechanisms. correlation to improve the prediction accuracy of station traffic in complex highway network environment.

[0058] In this embodiment, taking the traffic prediction of a highway site in a certain province as an example, the method is applied to the traffic prediction of a highway site. Specifically includes the following steps:

[0059] Step 1: Preprocessing of traffic data on expressway sites...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a method and a system for predicting the traffic of expressway sites based on a spatiotemporal attention mechanism. By constructing a training sample set of a target site, a site traffic prediction model based on a spatiotemporal attention mechanism is constructed, and training data in the training sample set is used for training. The site traffic prediction model enables it to learn the temporal dependence between the traffic data at the target site's predicted moment and its historical traffic data over a period of time in the past, as well as the relationship between the target site's traffic data at the predicted moment and its key source site's associated traffic data at the moment. and predicts the traffic data of the target site based on the temporal dependencies and dynamic spatial correlations obtained from training. The invention can capture the accurate dynamic spatial correlation between the traffic data of the target site and the traffic data of the key source sites, and combine the time dependence of the site traffic to predict the traffic, thereby improving the prediction of the site traffic in the complex expressway network. precision.

Description

technical field [0001] The invention relates to the field of traffic forecasting of intelligent transportation systems, in particular to a method and system for predicting traffic flow of expressway sites based on a spatiotemporal attention mechanism. Background technique [0002] Traffic flow forecasting is an important research field in intelligent transportation systems. Existing research methods mainly include methods based on statistics, traditional machine learning methods and methods based on deep learning models. Statistics-based methods are mainly designed for small data sets, are not suitable for processing complex and dynamic time series data, and ignore the spatial dependence of traffic data; traditional machine learning methods can process high-dimensional data and capture complex nonlinear relationships. However, its ability to mine complex spatiotemporal patterns is still limited and requires manual feature extraction by experts in the corresponding domain. ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06F30/20G06Q10/04G06Q50/26G06F119/12
CPCG06F30/20G06Q10/04G06Q50/26G06F2119/12
Inventor 吕丰王艺锋段思婧
Owner CENT SOUTH UNIV