Assessment of the safety impacts of travel demand management (TDM) policies must be conducted proactively. Because TDM policies are typically implemented at an aggregate level, crash prediction models should also be developed at a similar level. The resolution of these models should better match that at which evaluations of TDM policies are performed. Therefore, in this study zonal crash prediction models were considered to establish an association between observed crashes and a set of predictor variables in each zone. This analysis was performed with the generalized linear modeling procedure and the assumption of a negative binomial error distribution. Different exposure, network, and sociodemographic variables for 2,200 traffic analysis zones were considered predictors of crashes in the study area of Flanders, Belgium. An activity-based transportation model framework was applied to produce exposure measurements for crash data that consisted of injury crashes recorded between 2004 and 2007. Network and sociodemographic characteristics were also collected. Different zonal crash prediction models were developed to predict the number of injury crashes, including crashes involving fatalities and severe and slight injuries. These models were classified into three groups: (a) flow-based models, (b) trip-based models, and (c) a combination of the two. The results showed considerable improvement of model performance when both trip-based and flow-based exposure variables were used simultaneously in the formulation of the model. The main purpose of this study was to develop a predictive tool that could be used at the planning level to evaluate the impacts of different TDM policies on traffic safety.
Application of Different Exposure Measures in Development of Planning-Level Zonal Crash Prediction Models
Transportation Research Record
Transportation Research Record: Journal of the Transportation Research Board ; 2280 , 1 ; 145-153
2012-01-01
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Planning-Level Crash Prediction Models in Southern California
Transportation Research Record | 2022
|Macro and micro models for zonal crash prediction with application in hot zones identification
Online Contents | 2016
|