A case study of highway traffic flow model - model validation and simulation

In this paper, we present the traffic flow study of the section of Minnesota State Highway 194 between Arlington Avenue and Trinity Road, one of the most heavily traveled and congested roadways in Duluth, Minnesota. Due to the initial unsatisfactory results we found using the Papageorgiou's (1983) traffic flow model, we revised the model while retaining key logical elements. Based on the real data collected from the remote traffic microwave sensor based traffic detector, we identify the model parameters which give the best coincidence between the calibrated model and the real process. The model parameter identification procedure is formulated as an optimization problem which is solved by nonlinear programming. Our simulation results show reasonably good traffic flow estimation with acceptably small error on that road section.