Road condition prediction method and system based on real-time traffic flow and vehicle behaviors

CN119516783APending Publication Date: 2025-02-25浙江交投高速公路运营管理有限公司
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411659208.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-11-20
Publication Date
2025-02-25

AI Technical Summary

Technical Problem

Traditional traffic conditions prediction methods are based on historical data and statistical models, and the prediction accuracy is restricted by factors such as seasonal changes and emergencies, resulting in low prediction accuracy.

Method used

The road condition prediction method based on real-time traffic flow and vehicle behavior is adopted, combined with Internet of Things technology, big data analysis technology and artificial intelligence technology, vehicle behavior analysis is performed through the support vector machine model of the approximate nuclear method, and the vehicle flow analysis is performed by integrating spatiotemporal trajectory and density estimation. The tense evolution of road conditions is analyzed using an adaptive scaling dot product attention mechanism, and the road condition prediction model is optimized through the Newtonian conjugate gradient algorithm.

Benefits of technology

It improves the accuracy and efficiency of vehicle behavior analysis, enhances the comprehensiveness and accuracy of vehicle flow analysis, improves the accuracy and robustness of road condition prediction models, helps traffic management departments to timely understand road congestion, optimize traffic light control and road condition guidance, and improves traffic management efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119516783A_ABST
    Figure CN119516783A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of traffic condition prediction, and discloses a road condition prediction method based on real-time traffic flow and vehicle behaviors, and the method comprises the steps: collecting vehicle driving data in real time, the driving data comprises vehicle position, speed and acceleration information, and the vehicle position, the speed and the acceleration information are acquired; behavior analysis is carried out on the collected vehicle driving data to judge the vehicle running state; recognizing the density and mobility of road vehicles according to the vehicle behavior analysis result, and counting the road traffic flow to obtain the road traffic flow distribution condition; a road condition prediction model is constructed, and the model takes road traffic flow distribution time sequence information as input and takes congestion severity prediction results of different road areas as output; and carrying out optimization solution on the constructed road condition prediction model to obtain optimal model parameters, and instantiating the model according to the optimal model parameters to realize road condition prediction.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Cited By

  • Traffic signal optimization method based on model predictive control and trust region Bayesian optimization

    CN121459595A