Analysis on traffic flow data and extraction of nonlinear characteristic quantities

In the progress of urban traffic status evolution, we need the insights into the traffic performance which are very critical to the operation, control, maintenance, design and research of urban traffic. Many efficient approaches and analysis techniques are applied to analyze the macro and micro characterizes of traffic flow data. The attractors are reconstructed by making use of time-delay coordinates. Then the macro nonlinear characteristic quantities of urban traffic such as the Fractal dimension, Kolmogorov entropy and the largest Lyapunov exponent are extracted in this multi-dimension phase-space. The study on the temporal characteristics of these three parameters and the comparison between urban and freeway traffic flow data discovered that urban traffic flow is featured by some chaotic behaviors. The multi-fractal analysis of four typical modes of urban flow data manages to characterize and recognize the dynamic structure of urban traffic flow data more deeply and thus can be effectively exploited for the controlling and modeling of the network behaviors.