Dynamic estimation of freeway origin-destination demand and travel time using extended Kalman filtering algorithm

In the present research, a nonlinear Kalman filtering approach, i.e., extended Kalman filter (EKF) was proposed to solve dynamic OD flows and travel times on a freeway segment. The non-linearity results from the facts that the coefficient matrices in the measurement equation of the Kalman filtering framework are unknown in advance and needed to be obtained/updated in light of the most recent observations. The numerical results demonstrated the capability of the proposed EKF model in the dynamic estimation of freeway OD demands and travel times. More significantly, one can design beneficial traffic control and management strategies in accordance with the estimation results.