Online calibration-based express way traffic flow data filling method and system

Abstract

The present invention discloses an online calibration-based express way traffic flow data filling method and system. The method comprises a data acquisition step of grouping and numbering the detectors, and establishing a traffic flow database according to the traffic flow data acquired by the detectors; a real-time diagnosis step of diagnosing the traffic flow data acquired by the detectors real-timely, and diagnosing whether the data is lost or wrong; a model establishment step of establishing a multivariate linear model according to the time trend and the spatial relationship of the lane data of the groups of detectors; an online calibration step of establishing a parameter autoregression prediction model, and predicting the model parameters at the current moment based on the model parameters obtained according to the earlier stage data in the traffic flow database; a data filling step of filling the traffic flow data according to the multivariate linear model and the model parameters at the current moment. The online calibration-based express way traffic flow data filling method and the online calibration-based express way traffic flow data filling system of the present invention have the advantages of being simple to calculate, being small in storage space and accurate in prediction result, etc., and effectively make up the insufficiency of a static model method at the aspect of data filling.

Description

technical field [0001] The invention belongs to the field of intelligent traffic control systems, in particular to a method and system for filling expressway traffic flow data based on online calibration. Background technique [0002] In recent years, urban expressways have become more and more serious traffic congestion due to the increasing traffic demand. In order to alleviate congestion, intelligent traffic systems for expressway control have been applied in practice. Intelligent traffic control systems use traffic flow detectors to collect data, conduct traffic state estimation or prediction, and optimize and propose control decisions. Expressway control systems rely heavily on real-time traffic data, however, due to hardware failures or communication failures, real-time collected data are prone to data loss or errors. The effective operation of the traffic control system requires accurate and complete data. The loss of traffic data in a short period of time will caus...

AI Technical Summary

Problems solved by technology

However, these online imputation methods only consider one of the temporal and spatial relationships: replacing missing data with the average value of historical data at the same moment to reflect the daily variation trend of flow; or using spatial correlation with surrounding detectors to fill missing data to Reflect real-time fluctuations in traffic

However, only considering the temporal relationship cannot reflect the real-time characteristics of daily traffic flow, and only considering the spatial correlation cannot provide imputation results even when all surrounding detectors fail

There is no online imputation method that simultaneously combines temporal trends and spatial correlations to repair missing data

In addition, previous data imputation methods apply static model parameters, which may incorrectly predict the temporal and spatial variation of traffic flow

Images