Expressway traffic accident risk prediction method and related equipment

By introducing a residual network model of the multi-head attention mechanism in the prediction of highway traffic accident risk, the problem that traditional methods are difficult to capture long-distance dependencies is solved, the prediction accuracy and efficiency are improved, and more accurate spatial and temporal characteristic modeling is achieved.

CN119721339APending Publication Date: 2025-03-28XI'AN UNIVERSITY OF ARCHITECTURE AND TECHNOLOGY
5 Cites 1 Cited by

Patent Information

Application Number
CN202411771756.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-04
Publication Date
2025-03-28

AI Technical Summary

Technical Problem

Traditional highway traffic accident risk predictions are difficult to capture the dependence between sections that are far apart in spatial distribution but have similar accident risk characteristics.

Method used

A residual network model containing a multi-head attention mechanism is adopted, and the multi-source data of highway traffic accidents is preprocessed and feature selection, and the alternative variable set is constructed and standardized. The multi-head attention mechanism is used to enhance the model's ability to capture long-distance dependencies in the data.

Benefits of technology

It improves the accuracy and efficiency of highway traffic accident risk prediction, and can more effectively identify the dependence between sections with far-distance distances in spatial distribution but with similar accident risk characteristics, thereby improving the modeling ability of complex space-time characteristics.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses an expressway traffic accident risk prediction method, which belongs to the technical field of accident risk prediction, constructs model input containing various key features through preprocessing and analysis of multi-source heterogeneous data, comprehensively considers influence factors of expressway traffic accident risk prediction, and improves expressway traffic accident risk prediction accuracy. According to the method, a multi-head attention mechanism is introduced into the improved residual network, so that the capturing capability of the model on the long-distance dependency relationship in the highway traffic data is enhanced, and the prediction accuracy and the prediction efficiency are improved. The multi-head attention mechanism can identify the dependency relationship between road sections which are far away from each other in spatial distribution and have similar accident risk characteristics, so that the modeling capability of complex space-time characteristics is improved.
Need to check novelty before this filing date? Find Prior Art