An urban traffic accident risk prediction method based on road network

Abstract

A road network-based urban traffic accident risk prediction method relates to the technical field of traffic accident risk prediction, and solves the problems of temporal and spatial correlation and spatial heterogeneity. The process is: establishing a mapping relationship between accident locations and road sections; The similarity between the middle road sections is clustered to obtain the coarse-grained area and the accident risk of the coarse-grained area is calculated; the long-term features, short-term features and their corresponding weather features are fused in the time dimension, and spliced ​​after fusion; External feature E at time t t The attention mechanism is used to obtain the importance weight of each historical time slice; the weighted summation of the spliced ​​fusion data is obtained according to the weight to obtain the fusion result after the weighted summation; the fusion result and the output result of the shunting module are input into the feature layer, using The attention mechanism obtains the predicted accident risk value. The invention solves the problem of spatial heterogeneity, and takes into account the accuracy of prediction while dividing the space more finely.

Description

technical field [0001] The invention relates to the technical field of urban traffic prediction, in particular to a road network-based urban traffic accident risk prediction method. Background technique [0002] With the rapid development of urbanization, the rapid increase in the number of motor vehicles leads to frequent traffic accidents, causing casualties and huge economic losses. Therefore, predicting the risk of traffic accidents in the future has become a top priority. However, it is difficult to accurately predict the risk of traffic accidents, and its time distribution in days, weeks, and months is quite different, and complex factors such as crowd density, traffic flow, weather, and abnormal events will affect the accident risk. [0003] The early traditional machine learning methods mostly extracted road features: such as road shape, road speed, traffic flow on the road, etc., and used statistical models to perform regression analysis on the number of traffic acc...

AI Technical Summary

Problems solved by technology

[0006] In order to solve the problem of incomplete consideration of temporal-spatial correlation and spatial heterogeneity in existing methods, the present invention provides a road network-based urban traffic accident risk prediction method

Images