nemat soltani

Safety consideration in transportation planning requires macro-level safety indicators. This is possible using various variables in macro-level safety studies. Using macro variables lead to easier and less costly accident predictive models. In this study, the predictive model of accident frequency obtained using macro level variables. Independent variables include, total length of the road network, the ratio of the length of roads network across functional classification (principal arterial, minor arterial, collector and local) to the length of all the roads, ratio of length of the bus lines to the total length of road network and density of the intersections in a Traffic Analysis Zone (TAZ). Accordingly, information of 16137 accidents in 96 TAZ collected. Then, the TAZs prioritized based on safety using Empirical Bayesian Method and TAZs with the highest Potential for Improvement (PI) found. Based on the results, increasing in length of the road network and the ratio of the length of road network across minor arterial in a TAZ, increases the chance of an accident occurred in that TAZ and a higher ratio of collector roads and local roads decreases the number of accidents in the TAZ. In this regard, the ratio of collector roads has the least and the ratio of local roads has the greatest impact.

In this study, traffic and geometric factors affecting accidents occurring in road segments are investigated across different transportation modes (vehicle, motorcycle, and pedestrian) using micro and macro levels variables simultaneously while accounting for the effect of intra-zone correlation due to the same independent variables for accidents occurring within a zone. The data relating to 14903 accidents that had occurred in 96 Traffic Analysis Zones (TAZ) in Tehran were collected and imported into Geographic Information System (GIS) application. Negative Binomial models and multilevel models were adopted to predict the number of traffic accidents. Due to considering the multilevel structure of the data in multilevel models, it showed a better performance in explaining the factors affecting accidents. Moreover, based on the results obtained from analyzing the sensitivity analysis of variables for final models, the effect size of one variable in accidents varies across different modes of transport. This discloses the necessity of investigating accidents across modes of transport. According to the results, variables like the high intensity of intersection in one TAZ and the length of the road segment increase the number of traffic accidents in all three modes of transport. Variable of the ratio of the principal arterial length to total roads available in one zone has almost 3.2 times stronger effect on motorcycle accident than vehicle and pedestrian accidents. So that adding 1 unit to this variable increases the number of vehicle and pedestrian accidents by a factor of 1.7, whereas, this variable increases motorcycle accidents by a weight of 5.4.

Extensive studies have so far been carried out on developing safety models. Despite the extensive efforts made in identifying independent variables and methods for developing models, little research has been carried out in providing amendatory solutions for enhancing the level of safety. Thus, the present study first developed separate accident frequency prediction models by transportation modes, and then in the second phase, a development of safety improvement method (DSIM) was proposed.

To this end, the data related to 14,903 accidents in 96 traffic analysis zones in Tehran, Iran, were collected. To evaluate the effect of intra‑zone correlation, a multilevel model and a negative binomial (NB) model were developed based on both micro‑ and macro‑level independent variables. Next, the DSIM was provided, aiming at causing the least change in the area under study and with attention to the defined constraints and ideal gas molecular movement algorithm.

Based on a comparison of the goodness‑of‑fit measures for the multilevel model with those of the NB model, the multilevel models showed a better performance in explaining the factors affecting accidents. This is due to considering the multilevel structure of the data in such models. The final results were obtained after 200 iterations of the optimization algorithm. Thus, to decrease accidents by 30% and cause the least change in the area under study, the independent variable of “vehicle kilometer traveled per road segment” underwent a considerable change, while little change was observed for the other variables.

The final results of the DSIM showed that the ultimate solutions derived from this method can be different from the final solutions derived from the analysis of the results from the safety models. Hence, it is necessary to develop new methods to propose solutions for increasing safety.