Kernel Density Estimation of Traffic Accidents and Integrating Scale Analysis: a Case Study

Road safety is defined as the number of accidents occurs in a specific segment of a road. Thus, the frequency of accident happening can be a good indicator of road safety. As the frequency of accident decrease the road safety increases. Since accidents own spatial and temporal feature, finding ways to estimate these feature may be useful for traffic specialists take countermeasure decisions to prevent accidents from happening. Being able to estimate the density function of accidents may help the road specialist to have a better vision of road safety state. They can use proactive actions to prevent the accidents from happening. For instance, using the portable hazard signs to inform drivers of the danger. In this study, the accident distribution function for Qazvin-Karaj rural road is estimated with Kernel density estimation (KDE). Additionally, scale analysis is used to illustrate the impact of different bandwidth used for the purpose of road safety. The result of this paper illustrates the Accident prone locations (APLs) in this rural roadway and gives the road specialist an overview of the segments which have the highest risk of accidents to happen

کلیدواژه ها:

Road safety ، Accident distribution function ، KDE ، Scale analysis