Traffic flow prediction method and device based on attention mechanism, equipment and medium

By introducing an attention mechanism-based method in traffic flow prediction, using spatiotemporal encoder and multi-head self-attention mechanism to process traffic flow diagrams, the problem of inefficient acquisition of traffic flow prediction values ​​in the prior art is solved, and more efficient and reliable prediction results are achieved.

CN119741834AActive Publication Date: 2025-04-01湖南工商大学
9 Cites 0 Cited by

Patent Information

Application Number
CN202510228757.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-02-28
Publication Date
2025-04-01
Estimated Expiration
2045-02-28

AI Technical Summary

Technical Problem

The process of obtaining traffic flow prediction values ​​of the existing target traffic flow chart at the next moment is cumbersome, resulting in inefficient acquisition and susceptible to manual intervention.

Method used

The traffic flow prediction method based on the attention mechanism is adopted, and the preset traffic flow diagram and the enhanced preset traffic flow diagram are processed through the spatiotemporal encoder and the multi-head self-attention mechanism to generate the traffic flow prediction value of the target traffic flow diagram at the next moment.

Benefits of technology

The efficiency of obtaining traffic flow prediction value of the target traffic flow chart at the next moment is improved, the impact of manual intervention is reduced, and the reliability of the predicted value is improved.

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

Abstract

The invention relates to the field of intelligent traffic and the field of artificial intelligence, and discloses a traffic flow prediction method, device and equipment based on an attention mechanism and a medium, and the method comprises the steps: processing a preset traffic flow graph and an enhanced preset traffic flow graph through a space-time encoder and a multi-head self-attention mechanism in a traffic flow prediction model, obtaining a first positive sample pair, a first negative sample pair, a second positive sample pair and a second negative sample pair; splicing the feature vector of the first positive sample pair and the feature vector of the second positive sample pair to obtain a first spliced feature, and determining a third loss value based on the first spliced feature; adding the first loss value, the second loss value and the third loss value to obtain a target loss value; and selecting the traffic flow prediction model using the adjusted model parameters as a target model, and generating a traffic flow prediction value of the target traffic flow graph at the next moment through the target model. The prediction performance of the traffic flow prediction model can be improved.
Need to check novelty before this filing date? Find Prior Art