Method and system for forecasting regional traffic flow

Abstract

The invention relates to a regional traffic flow prediction method and system. The method includes: obtaining historical traffic flow data of vehicles in a target area, dividing the historical traffic flow data of the target area according to time periods, and obtaining historical traffic flow data of multiple cycles; selecting one of the historical traffic flow data of the cycle The data is input into the first traffic flow prediction model to obtain long-term and short-term traffic flow prediction data, and the historical traffic data of each cycle are respectively input into the second traffic flow prediction model to obtain a plurality of least squares traffic flow prediction data, and the plurality of all The least squares traffic flow forecast data is superimposed and reconstructed to obtain vector machine forecast data, and the long-term and short-term traffic flow forecast data and the vector machine forecast data are input into a weighted average forecast model to obtain regional traffic flow forecast data. The invention improves the prediction accuracy of regional traffic flow.

Description

technical field [0001] The invention relates to the technical field of space-time network flow prediction, in particular to a method and system for regional traffic flow prediction. Background technique [0002] The flow prediction problem of spatiotemporal network is a very challenging problem in intelligent transportation system. The nodes in the spatiotemporal network have spatial correlation according to the geographical topology, and the node traffic has temporal correlation in different time intervals. In the traditional neural network model in the prediction of spatiotemporal network traffic, due to the influence of the network structure and numerical calculation method, the prediction results have errors, and when the data scale is large, the time cost of model training is large, resulting in low training efficiency. SUMMARY OF THE INVENTION [0003] The purpose of the present invention is to provide a regional traffic flow prediction method and system, and the pr...

AI Technical Summary

Problems solved by technology

In terms of spatio-temporal network traffic prediction by traditional neural network models, due to the influence of network structure and numerical calculation methods, there are errors in the prediction results, and when the data scale is large, the time cost of model training is large, resulting in low training efficiency

Images