TOWARDS MODELING DRIVER BEHAVIOR UNDER EXTREME CONDITIONS

Abstract

The purpose of this study is to investigate the representation of driver behavior under extreme conditions, towards development of a micro-simulation modeling framework of traffic flow to support evaluation of management strategies and measures in emergency situations. To accomplish this objective, particular attention is given to understanding and representing "panic behavior" of individuals and how this behavior may be translated into driver actions. Related background from psychology and sociology is examined.

Following a systematic review of previous traffic models, a model is selected as a starting point for modification towards the micro-simulation of traffic flow under extreme conditions. The model is based on Gipps' (1981) Car-Following Model. To evaluate the proposed modification, a prototype implementation is proposed for the micro-simulation of traffic flow on a stretch of highway with simplified geometric features. The vehicle trajectories and aggregate traffic properties are evaluated with respect to different scenarios through a sensitivity analysis.