Road Safety Evaluation Based on BP Neural Network

Abstract

Evaluating road safety is essential in identifying the potential road safety hazard which could result in casualties and property losses. in this paper, a BP neural network was built by using neural network toolkit in "Matlab", Two similar roadways are used in calibrating and validating the network. The high level of predictability provided that the application of BP neural network model to road safety evaluation is feasible, and evaluation results coincide with the actual situation.