Impacts of car-following models on simulation-based safety evaluation of freeways

Abstract

This study investigates the sensitivity between car-following model selection, cooperative adaptive cruise-controlled vehicle penetration (CACC), and freeway traffic control on surrogate safety outcomes. We examine six calibrated car-following models, GM, GHR, Gipps, Krauss, Wiedemann-99, and IDM under three freeway traffic control strategies (No Control, ALINEA ramp metering, and an H infinity\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$H_\infty $$\end{document} integrated RM + VSL) and CACC market penetration rates from 0-100 %. Using time-to-collision and post-encroachment-time thresholds extracted from 108 microsimulation runs of an 11 km Istanbul corridor, we quantify how surrogate conflicts respond to each experimental factor. Results demonstrate that safety evaluations are inherently affected by the selected car-following model, and model sensitivity should be taken into account when claiming safety benefits of emerging vehicle technologies and control measures.