Verification and Calibration of Microscopic Traffic Simulation using Driver Behavior and Car Following Metrics for Freeway Segments
Typically, the validity of microscopic simulation scenarios is evaluated based on macroscopic measures such as travel time, delay, and queues. This is in part due to limitations or inability to collect field data at a microscopic scale (i.e., for an individual vehicle). The main objective of the proposed research is to leverage microscopic driving behavior and car-following metrics directly derived from Naturalistic Driving Study (NDS) datasets to develop guidelines to assist and enhance the calibration and verification of microscopic traffic simulation.
This project is intended to provide practitioners and researchers with new criteria to evaluate simulation from a microscopic point of view, complementing typical calibration efforts for macroscopic performance measures. The proposed systematic extraction of NDS driver behavior (such as free flow speed, acceleration, and deceleration rates) and car-following metrics (such as vehicle spacing and following speed), will provide unique ground truth data to better understand and characterize how drivers respond to different traffic conditions in freeway segments. The project team will use these metrics to calibrate simulated scenarios to enhance the simulation’s ability to reproduce traffic conditions and scenarios for planning, operational, and safety analysis.
- Safety Training and Analysis Center (STAC)
- Safety Data and Analysis
- Naturalistic Driving Study Pooled Fund