This report documents the analysis that produced detailed driving behavior distributions in terms of three main metrics: spacing between the instrumented vehicle and a leader in the same lane, acceleration of the instrumented vehicle, and acceleration change rate of the instrumented vehicle. The research team explored the potential benefits of incorporating measures derived from naturalistic data into traditional safety modeling. At a macroscopic scale, results from crash frequency prediction models showed consistent and significant effects associated with increases in multivehicle crash frequencies when the variance of the density increased, the variance of the speed increased, and the mean spacing decreased. The research team developed an open-source tool, the Naturalistic Assessments of Car-Following Trajectories tool, to implement the processes described in the report.

This research will be of interest to traffic operations and highway safety professionals, and others with an interest in mitigating disruptions to vehicle flow and improving safety.

Recommended citation: Federal Highway Administration, Verification and Calibration of Microscopic Traffic Simulation Using Driver Behavior and Car-Following Metrics for Freeway Segments (Washington, DC: 2024) https://doi.org/10.21949/1521775