Real-time freeway traffic state estimation based on extended Kalman filter: a general approach

Abstract

A general approach to the real-time estimation of the complete traffic state in freeway stretches is developed based on the extended Kalman filter. First, a general stochastic macroscopic traffic flow model of freeway stretches is presented, while some simple formulae are proposed to model real-time traffic measurements. Second, the macroscopic traffic flow model along with the measurement model is organized in a compact state-space form, based on which a traffic state estimator is designed by use of the extended-Kalman-filtering method. While constructing the traffic state estimator, special attention is paid to the handling of the boundary conditions and unknown parameters of the macroscopic traffic flow model. A number of simulations are conducted to test the designed traffic state estimator under various traffic situations in a freeway stretch with on/off-ramps and a long inter-detector distance. Some key issues are carefully investigated, including tracking capability of the traffic state estimator, comparison of various estimation schemes, evaluation of different detector configurations, significance of the on-line model parameter estimation, sensitivity of the traffic state estimator to the initial values of the estimated model parameters and to the related standard deviation values, and dynamic tracking of time-varying model parameters. The achieved simulation results are very promising for the subsequent development and testing work that is briefly outlined.