Safety Risk Factors Identification and Safety Risk Classification Assessment for Rural Roadsides

Identifying safety risk in rural roadsides is critical to reasonably allocated limited funds and avoid excessive installation of safety facilities. There is still a lack of abundant data for traffic accidents on rural roads in China, so scholars have screened the roadside safety risk factors by means of empirical qualitative methods. The existing assessment models are thus greatly restricted in practical applications. To overcome the above shortcomings, this paper first screens the roadside safety risk factors based on DEMATEL and ISM. Next, a rural roadside safety risk classification model is established by a Bayesian network. The model is capable of dealing with complex logical relations and inconsistent expert judgments, and is used for the quantitative classification of safety risk. Finally, the empirical analysis result shows that the proposed method possesses strong practicability when insufficient rural road accident data are available.