Traffic congestion propagation prediction method based on space-time diagram convolutional neural network

Abstract

The invention provides a traffic congestion propagation prediction method based on a space-time diagram convolutional neural network, and relates to the technical field of traffic big data mining andanalysis. Traffic data source data are acquired, a traffic speed data set and sensor network adjacent matrix data are made, a data driving mode is adopted, a space-time diagram convolutional neural network is introduced, and the traffic speed prediction precision of a target road section is improved. A speed prediction module is introduced into a congestion propagation model framework, so that thetraffic congestion prediction precision is improved, the traffic congestion propagation process is accurately represented, and the problems of low prediction precision, time-consuming algorithm operation and the like caused by poor spatial feature extraction of an existing traffic congestion propagation prediction scheme are solved; and the propagation condition of the congestion occurrence roadsection to the adjacent traffic road in the future time period is predicted.

Description

technical field [0001] The invention relates to the technical field of traffic big data mining and analysis, in particular to a traffic congestion propagation prediction method based on a spatiotemporal graph convolutional neural network. Background technique [0002] In today's society, many large cities regard traffic congestion as one of the problems that need to be solved urgently. When traffic congestion occurs in an area of ​​the traffic road network, this congested area will affect the traffic conditions of nearby roads, and even cause congestion on nearby sub-roads. Therefore, it is necessary to create an efficient model to predict the propagation of congestion in advance to predict when the adjacent road segments of the congestion source will be affected in the near future. In view of the above problems, the present invention proposes a traffic congestion propagation prediction method based on a spatiotemporal graph convolutional neural network, which has important...

AI Technical Summary

Problems solved by technology

XL.Ma and others built a deep convolutional neural network to predict short-term traffic flow speed, but it is difficult to accurately capture traffic information in the time domain

YP.Liu et al. proposed a convolutional long-short-term memory network to simultaneously capture the spatio-temporal characteristics of traffic speed to predict road traffic speed, but this model can only extract grid-type traffic data

However, the sensor network relies on the road network and is not a standard grid structure, so the performance of this predictive model is not good

In the actual traffic prediction scenario, due to the huge amount of data and the complexity of the traffic road network, it is difficult to capture the dynamic change of traffic speed in time and space, which leads to the poor prediction performance of the existing congestion propagation model.

Images