An Optimal Weight Based Multi-parameter Fusion Clustering Method for Highway Traffic Safety State Division

Abstract

In order to deeply analyze the quantitative relationship between traffic flow state and crash risk, this paper presents an optimal weight based multi-parameter fusion clustering method for highway traffic safety state division. Firstly, highway crash data and the corresponding upstream and downstream traffic flow data were extracted and matched with the paired case-control method. Secondly, considering the different roles of the three parameters of traffic volume, speed, and occupancy in the classification of traffic state, the weight optimization algorithm is applied to calculate the three parameters weight in this study. Therefore, the traffic state comprehensive evaluation index with three parameters fusion is obtained and taken as the input index of traffic safety state division. Finally, the k-means clustering method is used to partition the highway traffic safety state. The results of case study show that the method proposed in this paper can realize the rational and effective division of traffic safety state. The classification results are helpful to quantitatively evaluate the highway crash risk levels under different traffic safety states. Furthermore, according to different traffic safety state division results, differentiated preventive measures could be formulated to improve highway traffic safety environment.