Calibration of Vissim Models for Rural Freeway Lane Closures: Novel Approach to the Modification of Key Parameters

Abstract

The definition of freeway work zone capacity has been a topic of debate for several decades, leaving agencies with limited guidance on predicting the behavior of traffic flow at given volumes for various work zone configurations. The methodology presented in the recently published 6th edition of the Highway Capacity Manual (HCM) is a substantial improvement over historical guidance and provides estimates of the mean queue discharge rate under a variety of prevailing site conditions. However, it is limited by the fact that its outputs are deterministic, while traffic flow and breakdown are stochastic phenomena. Recently, well-calibrated microsimulation models have shown promise as a freeway work zone traffic analysis tool, but most guidance is focused on site-specific modeling. This research aimed to address these shortcomings by presenting a novel approach to developing and calibrating generalizable microsimulation models for rural freeway lane closures in Vissim, a traffic simulation software package developed by the PTV Group. Specifically, it was determined that such models may best replicate field conditions at rural freeway work zones when time headway is described by a field-measured distribution and truck characteristics are representative of the United States (U.S.) fleet. The results suggested that the default desired acceleration for heavy trucks should be set between 2 and 3 ft/s2 and that separate time headway distributions should be constructed for passenger cars and trucks. The methodology presented herein may be extended to obtain stochastic estimates of capacity for sites exhibiting a variety of geometric, traffic, and environmental characteristics.