Presenting a prediction model for suburban roads accidents

Road accidents are one of the most important causes of death in the world,especially in Iran. . Annually, many people in the world are killed and injuredas a result of accidents caused by human error on the suburban roads. Giventhe increasing population growth and the increase in road accidents in thecountry, finding the most effective accident factors can lead to a reduction indamages and casualties resulting from these accidents and a proper basis forroads accident management practices. Therefore, the purpose of this study wasto identify the factors affecting the incidence of accidents resulting humanerror leading to injury or death, as well as to present a model for predictingaccidents due to the impact of human error in suburban roads of Hamadanprovince during 2012-2017. To collect the data, the data collected were usedusing the accidents unprepared form. From 1345 accidents registered in thecase study, 940 accidents were selected in the incident parts, out of which 610accidents due to carelessness in 97 plots were selected for use in the model. Inthe modeling process, were used accident data, traffic and geometry of theincident parts of the case study that the geometrical data of the identifiedincident parts were collected as field and using two-frequency GPS. Findingsshowed that Poisson model is the best model in terms of accident predictionand ability to present formula. Five factors were effective in this model that theindicator factor of proximity to population centers with the coefficient of1/13081 had the most effect in car accidents caused by carelessness and factorssuch as slope and curvature composition, crash hour traffic, segmentationlength and slope are next hierarchy. The results of this study can be a properbasis for management disaster management practices caused by human error.Also the factors affecting accidents, will be used as a suitable tool at themiddle level of road safety management.