Traffic flow three parameter real time prediction method taking regard of space-time correlation

Abstract

The invention discloses a traffic flow three parameter real time prediction method taking regard of space-time correlation. According to the method, on the basis of acquiring traffic flow rate, speed and occupancy data of a target section and upstream and downstream sections of the target section, a state space model for multivariable short time prediction of traffic flow three parameters is established; according to spatial correlation of various traffic variables at different data acquisition sections, an observation equation of the state space model is established; according to time autocorrelation and cross correlation of the multiple traffic variables at one same data acquisition section, a state equation of the state space model is established; prediction and iteration update of the traffic flow three parameters are realized by employing the Kalman filtering algorithm. The method makes full use of the spatial correlation of the traffic flow three parameters at the different data acquisition sections, the time autocorrelation and the cross correlation of the different traffic variables at one same data acquisition section, the multivariable prediction algorithm is employed, and thereby traffic flow short time prediction accuracy is facilitated.

Description

technical field [0001] The invention relates to the field of traffic forecasting, in particular to a real-time forecasting method for three parameters of traffic flow considering temporal and spatial correlation. Background technique [0002] Traffic flow short-term forecasting technology is an important research content in the field of intelligent transportation. Accurate and reliable real-time prediction of traffic conditions is an important data support for the realization of advanced traffic information services, active traffic guidance, active traffic signal control and many other intelligent traffic operation management and control. [0003] At present, a lot of research has been carried out on the real-time prediction technology of traffic conditions at home and abroad. Forecasting methods based on statistical models and artificial intelligence technologies have been continuously proposed, and the accuracy of predictions has also been continuously improved. However, t...

AI Technical Summary

Problems solved by technology

[0003] At present, a lot of research has been carried out on the real-time prediction technology of traffic conditions at home and abroad. Forecasting methods based on statistical models and artificial intelligence technologies have been continuously proposed, and the accuracy of predictions has also been continuously improved. However, the input and The output is mainly based on the data of a single detection section or a single traffic variable. When studying the real-time prediction of traffic conditions of continuous flow facilities such as expressways and urban expressways, there is a lack of spatial correlation between continuous section traffic flows and the same section. Utilization of effective information such as time correlation between different traffic variables

Images