Modeling and Predicting the Number of Suburban Road Accidents
Through ever-increasing economic growth, road transportation has played an essential role in State Transportation Comprehensive System. By increasing road transport in Iran, fatalities have remarkably increased. Thus, the necessity of planning for reduction of destructive effects and measuring of effectiveness for the previous conducted activities will be augmented. Measurement and planning will not be realized without gathering information and data concerning to history and prediction of the future status. The complexity, uncertainty of the factors for occurrence of accidents, and limitation of documented data about quantity of road accidents correspond to the characteristics of Theory of Grey Systems. GreyModels (GMs) requires a restricted volume of data for estimation of behavior of unknown systems. In the present paper, data from city outside road accidents (2012-2006) have been utilized to evaluate the accuracy of prediction models of GM (1,1), DGM(1,1), DGM(2,1),RGM(1,1) FGM (1,1), and Verhulst Model. The used data are classified into four groups i.e. accidents with damages, injuries, fatalities, and total accidents. The average minimum error serves as evaluation criterion for accuracy of models. In this technique, the mean error of the predicted series and series of primary data are computed. Then, number of occurred accidents in year 2013 and 2014 will be predicted by means of DGM (1,1), which has had the maximum accuracy rate in this prediction