California Partners for Advanced Transportation Technology

This report documents the procedures the PATH team has implemented for calibrating the NGSIM oversaturated flow model in the Aimsun micro simulation environment. The goal of the calibration is to identify suitable driving behavior parameters of the NGSIM oversaturated flowmodel, which is used to describe the behaviors of human drivers. With these parameters, the model should be able to reproduce traffic flow patterns like those observed in the 18-km State Route 99 corridor to the south of Sacramento, CA. The Geoffrey E. Havers (GEH) statistic is adopted to quantify the relative difference between the simulated and observed flows. The calibrated model is considered satisfactory if the GEH statistic is less than 5 in at least 85% of the time intervals.  In addition to the GEH statistic, the simulated traffic pattern must produce spatial and temporal speed distribution and flow-density relationships similar to those observed in the field.