Recent years have seen major advances in the field of simulation-based dynamic traffic assignment (DTA), resulting in the development of DTA software packages capable of simulating real-world networks. However, simulation of such networks requires not only sophisticated algorithms and software, but also large and detailed data sets. Although algorithmic and software-related issues in large-scale DTA development have received considerable research attention, there is little reported experience with the data-related challenges in real-world applications of large-scale simulation models. There were challenges in using a large-scale simulation-assignment model for evaluation of transit signal priority (TSP) in the Chicago, Illinois, region. Relevant impacts of TSP are described, to provide a framework for comparing simulation approaches and data sets. The practice of using microsimulation models to evaluate TSP impacts on short corridors is compared with that of using regional assignment-simulation approaches, with an emphasis on TSP impacts that can be captured and observed with each one of the approaches. The data sets used for the regional Chicago TSP study are then described, along with assumptions made to adapt each data set to the task of regional time-dependent simulation.
Multiple rotations of Gaussian quadratures: An efficient method for uncertainty analyses in large-scale simulation models
