Urban traffic accident risk prediction method based on road network

Abstract

An urban traffic accident risk prediction method based on a road network relates to the technical field of traffic accident risk prediction, solves the problems of spatial-temporal correlation and spatial heterogeneity, and comprises the following steps: establishing a mapping relation between an accident position and a road section; clustering the road sections according to the similarity between the road sections in the road section set to obtain a coarse-grained area, and calculating the accident risk of the coarse-grained area; fusing the long-term features, the short-term features and the corresponding weather features in the time dimension, and carrying out splicing after fusion; according to the external feature Et of the to-be-predicted t moment, adopting an attention mechanism to obtain an importance weight in each historical time slice; performing weighted summation on the spliced fusion data according to the weight to obtain a fusion result after weighted summation; and inputting a fusion result and an output result of the shunting module into a feature layer, and obtaining a predicted accident risk value by adopting an attention mechanism. According to the method, the problem of spatial heterogeneity is solved, and the prediction accuracy is considered while spatial division is finer.

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 and the rapid increase in the number of motor vehicles, traffic accidents occur frequently, 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. The time distribution varies greatly between days, weeks, and months. Complex factors such as crowd density, traffic flow, weather, and abnormal events will all affect the accident risk. [0003] Most of the early traditional machine learning methods extracted road features: such as road shape, road speed, traffic volume on the road, etc., and used statistical models to perform regression analysis on the number of traffic ...

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