Spatial Accident Prediction Modeling: A Case Study of Hyderabad City

Abstract

The main objective of this paper is to develop the spatial accident prediction model by considering population, resident worker, resident student, employment, enrolled student, road density and trips generated. Road traffic accident data of Hyderabad city is collected from 2011 to 2015. Total Accidents per year and Total persons injuries per year are considered as independent variables. In this study, 263 traffic analysis zones are considered in Hyderabad city. Spatial Analysis carried out to extract zonal wise dependent and independent data by using QGIS software. Blackspot was found for each TAZ and area, length of the road network, the number of accident location in each TAZ has been found using QGIS. Multiple Linear Regression is used to develop the relationship between the independent and dependent variables. Total four different accident prediction models are developed, in which, model 1 and 2 deal with total accidents per year as dependent variable while total persons injured per year in model 3 and 4. Results are concluded that population density and resident student density are showing the negative impact to model while other parameters are showing a positive relationship with the dependent variable. When population density increase then the population in that zone also increases but accident decreases. Generally, people are living near to their workplace to reduce the daily trips and save travel time and cost. So, if population density increases then the person taking a short trip and with trips having a positive impact on accident, it can be concluded that with increases in population density accident will reduce. Similarly for resident student also, because students are living near to their school, college or institutes hence total trips
generated by the student will decrease. But some student who enrolled in a particular zone living far from there, it generated longer trips hence positive impact on accident. Road density increases tend to increase the number of intersections, bus stop, bridge hence lead to an increase in black spot and more black spot more accident will occur.