Road condition assessment method of short-term traffic prediction map convolutional network

Abstract

The invention relates to a road condition evaluation method of a short-term traffic prediction map convolutional network, and belongs to the field of intelligent traffic. An adjacent matrix representing a distance binary relationship of nodes in a traffic network is abstracted into a fuzzy relationship matrix, a transfer relationship matrix representing a relationship among all potential nodes is obtained by calculating a transitive closure of the fuzzy relationship matrix, and a transfer relationship matrix based on current input is obtained by combining the transfer relationship matrix and a self-learning matrix of an attention mechanism, and the transfer relationship matrix based on current input is used for a graph convolution process. According to the method, the transfer relationship of the data in the traffic network is represented through weight learning, and the capability of capturing the dynamic space dependence in the traffic data flow by the graph convolutional network is improved. According to the model, three types of data sets are constructed, the long time span dependence in the traffic flow is considered; causal cavity convolution is employed, the short time span is considered; and therefore, the accuracy of traffic data prediction can be improved.

Description

technical field [0001] The invention relates to a road condition evaluation method of a short-term traffic prediction graph convolution network, belonging to the technical fields of intelligent transportation and artificial intelligence. Background technique [0002] Against the background of the substantial progress in the level of the automobile industry and the increasing economic level of residents, the number of cars owned by residents in my country continues to increase, resulting in increasingly serious problems such as traffic congestion and environmental pollution. The development of a green and efficient intelligent transportation system is a national strategy and the aspiration of the people. Traffic flow is an important data for intelligent transportation system analysis, because the traffic flow in a region can well reflect the current traffic conditions in the region. Using historical traffic flow data to analyze the potential laws of traffic flow and make pre...

AI Technical Summary

Problems solved by technology

However, the spatio-temporal relationship may change dynamically, and different road sections may have different influences on each other at different time nodes. This dynamic feature is usually not effectively captured.

Images