Modeling traffic flow correlation using DFA and DCCA

The focus of the present paper is on the power-law auto-correlations and crosscorrelations in traffic time series. Detrended fluctuation analysis (DFA) and detrended cross-correlation analysis (DCCA) are used to study the traffic flow fluctuations. We find that the original traffic fluctuation time...

Full description

Saved in:
Bibliographic Details
Published in:Nonlinear dynamics 2010-07, Vol.61 (1-2), p.207-216
Main Authors: Xu, Na, Shang, Pengjian, Kamae, Santi
Format: Article
Language:English
Subjects:
Autocorrelation
Automotive Engineering
Classical Mechanics
Control
Correlation
Correlation analysis
Dynamical Systems
Engineering
Fluctuation
Lanes
Mathematical models
Mechanical Engineering
Original Paper
Paths
Power law
Time series
Traffic engineering
Traffic flow
Traffic models
Traffic speed
Variation
Vibration
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The focus of the present paper is on the power-law auto-correlations and crosscorrelations in traffic time series. Detrended fluctuation analysis (DFA) and detrended cross-correlation analysis (DCCA) are used to study the traffic flow fluctuations. We find that the original traffic fluctuation time series may exhibit power-law auto-correlations; however, the sign-separated traffic fluctuation signals, both the positive fluctuation signals and the negative fluctuation signals, exhibit anti-correlated behavior. Further, we show that two original traffic speed fluctuation time series derived from adjacent sections exhibit much stronger power-law cross-correlations than the two time series derived from adjacent lanes. Finally, we demonstrate that for two sign-separated traffic fluctuation signals, there exist long-range cross-correlations between the positive fluctuation signals and the negative fluctuation signals, derived from two adjacent lanes, respectively. But, for two same-sign traffic fluctuation signals derived from two adjacent lanes, there is power-law cross-anti-correlation in the variables.
ISSN:0924-090X
1573-269X
DOI:10.1007/s11071-009-9642-5