Graph convolutional neural network based urban traffic flow prediction method and medium

Abstract

The invention discloses a graph convolutional neural network based urban traffic flow prediction method and a medium. The method comprises that original data is obtained; a distance matrix is generated according to latitude and longitude information corresponding to each node; a reachability matrix is calculated according to a mean value of the speed limit and the distance matrix; an initial traffic flow prediction model for predicting the traffic flow velocity is established, traffic flow velocity information and the reachability matrix are input to the initial traffic flow prediction model,so that the initial traffic flow prediction model outputs a traffic flow velocity predicted value according to the traffic flow velocity information and the reachability matrix; the initial traffic flow prediction model is trained to determine a final traffic flow prediction model; and traffic flow velocity information to be predicted and the reachability matrix to be predicted are input to the traffic flow prediction model, so that traffic flow in the future is predicted via the traffic flow prediction model. Thus, spatial characteristics of the urban traffic road network are effectively extracted, the traffic flow is predicted more accurately, and the prediction method is more universal and convenient to popularize.

Description

technical field [0001] The present invention relates to the technical field of information processing, in particular to a graph convolutional neural network-based urban traffic flow prediction method and a computer-readable storage medium. Background technique [0002] Traffic flow refers to the traffic flow formed by continuous driving of cars on the road. With the increasing number of cars, urban traffic is becoming more and more congested. Therefore, it is particularly important to predict urban traffic flow. For example, through traffic flow prediction, it is convenient for relevant departments to control urban traffic in a directional way; Travel time of motorists, etc. [0003] However, in related technologies, it is difficult to effectively obtain the unique physical characteristics of the traffic road network, resulting in low accuracy of the final predicted traffic flow data; moreover, the requirement for the amount of information of the original data is too high (...

AI Technical Summary

Problems solved by technology

[0003] However, in related technologies, it is difficult to effectively obtain the unique physical characteristics of the traffic road network, resulting in low accuracy of the final predicted traffic flow data; moreover, the requirement for the amount of information of the original data is too high (for example, the adjacency information of sensor nodes etc.), lack of universality

Images