Severe Crash Risk Distribution Model between Each Origin-Destination Pair of Traffic Analysis Zones- The Case Study of the City of Qom in Iran

Crash prediction models (CPMs) are a valuable tool in transportation planning studies that enable safety and its changes to be investigated in future decisions. The main goal of this study is to develop the CPMs based on the trip distribution step from the common four-step demand models, which is called the crash risk distribution in this study. In order to, the Negative Binomial formulation with a log-link function is used. The frequency of severe crash in the city of Qom, Iran, including the total number of fatal and injury crash among each origin-destination (OD) pair of traffic analysis zones (TAZs) as a dependent variable, and explanation variables, including the traffic-related and trip distribution results, have been used for model calibration. Traffic-related data include the vehicle kilometers traveled, experienced travel time with vehicle, and average vehicle speed between each OD pair of TAZs. The results of trip distribution include the trip distribution by purposes of work, education, shop, and personal. Model calibration and validation results show a significant relationship between the mentioned variables and the frequency of severe crashes. Hence, the developed model predicts the number of severe crashes in future years based on increased trips distributed among TAZs. Therefore, with the help of this model, it is possible to identify and prioritize trips with high-risk origin-destination in terms of safety, and the impact of different travel demand management scenarios on safety can be evaluated.