查看文章

Real-time traffic flow models can be used for traffic management in networks or the estimation or prediction of the traffic state. In urban environments, intersection delays pose a problem to the macroscopic traffic flow models, while a macroscopic approach is desirable due to the calculation speed advantages of these models on larger networks. Current intersection models generally do not sufficiently take delays due to crossing traffic into account. In this paper, a Linearly Decreasing Turn Capacity (LDTC) is introduced to macroscopically model intersection delays. In JDSMART, the model that is used, the Godunov scheme is adapted to incorporate the turn capacities. Along with validating the linear form of the LDTC model, a procedure is described to calibrate its parameters based on simulated or actual traffic data. Finally, the approach is applied to an urban test network, which is also modelled in the microscopic …