Freeway traffic safety state classification method based on multi-parameter fusion clustering

Abstract

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.