Power-law fluctuation in expressway traffic flow: Detrended fluctuation analysis

摘要

The temporal behavior of expressway traffic flow is a complex mixture of various time scales. The fundamental response time of drivers is on the order of 1 s, and the periodic appearances of traffic congestion reflect Our social activities. We need some filters to extract proper dynamical fluctuations from raw observed traffic data. We employ the detrended fluctuation analysis (DFA) for studying long-range correlation in the time sequence of traffic flow. The time series of expressway traffic flow is understood to consist of periodic trends with power-law fluctuation. The power spectrum P(k) behaves as P(k) similar to k(-beta) with beta similar to 1. The power-law fluctuation is observed in the wide time range from minutes to months.