A Method for Forecasting Regional Radial Traffic Flow

Abstract

The invention discloses a regional radial traffic flow prediction method, which is characterized in that it includes the following steps: S1: collecting trajectory data information of residents' movement in daily life, and converting the trajectory data into traffic entry and exit with sub-region as the smallest unit Flow map; S2: Obtain the characteristics of traffic flow in different areas of the city, and the spatial correlation of traffic between different areas of the city; S3: Build a regional radial traffic flow prediction model, using a unique full convolutional network starting from the central area, By superimposing feature extraction-prediction sub-modules to realize gradual expansion prediction; S4: By adjusting the full convolutional neural network inversely by adjusting the error between the predicted flow map and the actual flow map, and fitting the regional radial traffic close to the real situation traffic forecasting model. The invention solves the problem of continuously extending the forecast based on the traffic flow in the central area to the surrounding area, shortens the time period of the forecast, and realizes the real-time forecast of the traffic flow.

Description

technical field [0001] The invention relates to the field of traffic flow forecasting, in particular to a regional radial traffic flow forecasting method. Background technique [0002] With the development of the economy, the number of private cars continues to increase, and the method of strengthening road infrastructure construction and increasing road traffic capacity has been unable to meet the growing traffic demand. Traffic congestion, traffic safety hazards and lack of traffic information not only bring inconvenience to travelers, but also bring resistance to urban development. The real-time grasp and accurate prediction of regional traffic flow become an important basis for solving traffic problems by using intelligent transportation technology (ITS). [0003] Most of the methods for traffic forecasting are based on time series. This type of method requires that traffic must have a great relationship with time, and the time series must be largely seasonal. The time...

AI Technical Summary

Problems solved by technology

[0005] The object of the present invention is to provide a regional radial traffic flow forecasting method for the above-mentioned existing problems, which solves the problem of continuous expansion forecast based on the flow rate in the central area to the surrounding area, and shortens the time period of forecasting

Images