Traffic jam prediction method for overpass based on data drive

Abstract

The invention discloses a traffic jam prediction method for an overpass based on data drive, and the method comprises the following steps: collecting the traffic state data of entrance and exit rampsof an overpass road surface; carrying out the preprocessing of the traffic data of the exit ramp of the overpass, and obtaining the vehicle flow speed and density of a controlled road segment; constructing a macroscopic traffic flow map of the controlled road segment through three indexes (traffic flow, speed and density), and calibrating a jam speed threshold value; selecting the speed as a jam feature, extracting a historical jam speed sequence, carrying out the clustering according to the difference of jam degree and obtaining the jam type; constructing a multi-scale speed predictor according to the historical speed data; inputting a real-time speed data sequence into the multi-scale speed predictor, obtaining a predicted speed sequence, and carrying out the jam prediction through a classifier. The method can achieve the effective prediction of the traffic situation in a certain future period according to the historical traffic situation, achieves the judgment and piling of the predicted speed sequence, and achieves the effective judgment of the traffic jam degree in future.

Description

technical field [0001] The invention relates to a data-driven viaduct traffic jam prediction method, more specifically a method for fitting traffic rules based on historical data and predicting future traffic jams. Background technique [0002] In recent years, more and more attention has been paid to the research on using large amounts of data to mine behavior rules in the traffic field and assist management decision-making. The development of big data technology has promoted the development of the field of intelligent transportation. Using long-term accumulated data, we can dig out the operation rules of the traffic system, and then make predictions, judgments, and decisions on future traffic conditions. One of the most concerned issues in the traffic field is whether there is congestion in the traffic situation and how to relieve the congestion. However, it is often difficult to dredge after traffic jams occur. Therefore, the present invention proposes a data-driven vi...

AI Technical Summary

Problems solved by technology

However, in the above methods, only the historical data of a few moments before the forecast time are considered, and the periodicity of the data with a larger time scale is not considered, so the accuracy of the model is still lacking

Images