In order to quantitatively analyze the influence of different traffic conditions on highway crash risk, a method of crash risk assessment based on traffic safety state division is proposed in this paper. Firstly, the highway crash data and corresponding traffic data of upstream and downstream are extracted and processed by using the matched case-control method to exclude the influence of other factors on the model. Secondly, considering the weight of traffic volume, speed and occupancy, a multi-parameter fusion cluster method is applied to divide traffic safety state. In addition, the quantitative relationship between different traffic states and highway crash risk is analyzed by using Bayesian conditional logistic regression model. Finally, the results of case study show that different traffic safety conditions are in different crash risk levels. The highway traffic management department can improve the safety risk management level by focusing on the prevention and control of highrisk traffic safety conditions.
In order to deeply analyze the quantitative relationship between traffic flow state and crash risk, a highway traffic safety state classification method based on multi-parameter fusion clustering was proposed. First, attribute data of highway traffic crashes and corresponding upstream and downstream traffic flow data were extracted, and matched with paired case-control method. Secondly, considering the different roles of traffic volume, speed and occupancy in traffic state classification, the weight optimization algorithm is introduced to calculate the weight of the three parameters. Therefore, the comprehensive evaluation index of traffic state with the fusion of three parameters is obtained and used as the input index of traffic safety state clustering. Finally, [Formula: see text]-means clustering method is used to classify the highway traffic safety status. The result of the case study shows that the proposed method can achieve reasonable and effective traffic safety state division. The classification results are helpful to quantitatively evaluate highway crash risk levels under different traffic safety states.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.