Verification and Calibration of Microscopic Traffic Simulation Using Driver Behavior and Car-Following Metrics for Freeway Segments

Medina, Juan C.; Malekloo, Arman; Kersavage, Kristin; Porter, Richard J.; Liu, Xiaoyue Cathy · 2024 · ROSA P / United States. Federal Highway Administration. Office of Safety and Operations Research and Development

archive: archived pipeline: cataloged verified

Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)

Summary

This research addresses the limitation of traditional microscopic traffic simulation calibration, which typically relies on macroscopic metrics like travel time and delay rather than individual vehicle interactions. The study aims to develop a calibration process that ensures simulated vehicle-to-vehicle interactions reflect naturalistic driver behavior, thereby improving the accuracy of traffic simulations for safety analysis and operational planning. The work was motivated by the availability of the Second Strategic Highway Research Program (SHRP2) Naturalistic Driving Study (NDS) datasets, which provide detailed real-world driving data. The researchers utilized over 1,600 hours of car-following data from more than 1,700 unique drivers across freeway segments in Florida, North Carolina, and Washington. Data collection was conducted in two phases: Part A used readily available datasets to identify suitable metrics, while Part B expanded the dataset to cover a wider range of traffic conditions, particularly lower-speed scenarios. The analysis focused on three primary microscopic metrics: vehicle spacing, acceleration, and acceleration change rate (jerk). These metrics were categorized by speed groups ranging from 5 mph to 85 mph to create detailed driving behavior distribution targets. To facilitate the calibration process, the team developed the Naturalistic Assessments of Car-Following Trajectories (NACT) tool, an open-source application that extracts car-following behavior from simulation outputs and performs statistical comparisons against the NDS targets. The study found that incorporating these microscopic NDS-derived metrics into safety modeling yielded significant results. At a macroscopic scale, crash frequency prediction models indicated that increases in multivehicle crash frequencies were significantly associated with increased variance in traffic density, increased variance in speed, and decreased mean vehicle spacing. At the microscopic level, simulations calibrated using the NDS targets produced vehicle conflict frequencies and locations that more closely resembled observed crash events compared to uncalibrated simulations. This suggests that microscopic calibration enhances the validity of surrogate safety assessments, particularly in low-traffic conditions. The significance of this work lies in providing a model-independent framework for calibrating microscopic traffic simulations using naturalistic driving data. By complementing traditional macroscopic calibration with microscopic behavioral targets, the proposed process improves the simulation's ability to emulate real-world vehicle interactions. The development of the NACT tool offers practitioners a practical method to verify and calibrate simulations, potentially leading to more accurate safety evaluations and better mitigation of traffic flow disruptions. The authors recommend further research to evaluate these benefits across a comprehensive set of traffic conditions, including high-traffic scenarios and high-demand fluctuations.

Key finding

Calibrating microscopic traffic simulation using naturalistic car-following metrics for spacing, acceleration, and jerk improves the representation of vehicle-to-vehicle interactions and enhances the accuracy of surrogate safety analysis.

Methodology

naturalistic

Sample size: 1700

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• following distance
• traffic density
• naturalistic crash near crash
• speed choice
• exposure measurement

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Methodological Resource: dataset resource, validation psychometrics
• Theoretical Contribution: computational model