Comparison of Low-Fidelity TRANSIMS and High-Fidelity CORSIM Highway Simulation Models with Intelligent Transportation System Data

In recent years there has been increased emphasis in the transportation modeling field on replacing macroscopic supply functions with simulation models. For example, the highway supply relationship in the Transportation Analysis and Simulation System (TRANSIMS) is based on a low-fidelity microsimulation model. How the TRANSIMS low-fidelity highway simulation module compares with a high-fidelity model and with empirical observations from intelligent transportation system (ITS) implementation projects has been examined. A section of Interstate 10 in Houston, Texas, was used as a test bed and ITS data were obtained for calibration and validation purposes. For comparison, the high-fidelity CORSIM model, which is used extensively in North America for operational analyses, was calibrated and tested with the same data. The two models did equally well at replicating the baseline volume data. In addition, the mean travel time output from the calibrated TRANSIMS model tended to be about 20 percent greater than the mean travel time from the calibrated CORSIM model. In general, the observed travel times were found to lie between the simulated values from the TRANSIMS and CORSIM models. More important, the link and corridor travel time variability appeared to be significantly less than the observed travel time variability. It is hypothesized that this difference may affect certain measures of effectiveness, such as automobile emissions, that will be used by transportation planners.