Abstract
With the development of freeway system informatization, it is easier to obtain the traffic flow data of freeway, which are widely used to study the relationship between traffic flow state and traffic safety. However, as the development degree of the freeway system is different in different regions, the sample size of traffic data collected in some regions is insufficient, and the precision of data is relatively low. In order to study the influence of limited data on the real-time freeway traffic crash risk modeling, three data sets including high precision data, small sample data, and low precision data were considered. Firstly, Bayesian Logistic regression was used to identify and predict the risk of three data sets. Secondly, based on the Bayesian updating method, the migration test towards high and low precision data sets was established. Finally, the applicability of machine learning and statistical methods to low precision data set was compared. The results show that the prediction performance of Bayesian Logistic regression improves with the increasing of sample size. Bayesian Logistic regression can identify various significant risk factors when data sets are of different precision. Comparatively, the prediction performance of the support vector machine is better than that of Bayesian Logistic. In addition, Bayesian updating method can improve the prediction performance of the transplanted model.
Highlights
In recent years, the potential safety hazard of new energy vehicles has gradually attracted attention, especially the accident of pure electric vehicles [1, 2]
E receiver operating characteristic (ROC) curve can be drawn by using true positive rate (TPR) and false-positive rate (FPR). e ROC represents the curve of the prediction accuracy of the data set under different probability thresholds. e AUC value of the area under the ROC curve can be calculated to measure the quality of the model. e closer the AUC value is to 1, the better the performance of the model
Considering the influence of limited data conditions on the real-time freeway traffic crash risk model, this paper constructed high precision data set, low precision data set, and small sample data set. ese data sets were modeled and analyzed based on Bayesian Logistic regression, and the reliability of real-time crash risk model transplantation based on Bayesian update was verified
Summary
The potential safety hazard of new energy vehicles has gradually attracted attention, especially the accident of pure electric vehicles [1, 2]. Tian et al analyzed the temporal and spatial distribution characteristics of freeway crashes in mountainous areas based on historical crash data, identified the significant influencing factors, and proposed corresponding improvement strategies [5]. Based on loop detector data and crash data collected by the Shanghai expressway system, Sun et al established a Bayesian network (BN) model to analyze real-time traffic flow parameters and crash risk of expressway [27]. Ma et al established a crash risk assessment and analysis model using highway crash data and real-time traffic flow data. From the perspective of data, this paper uses statistical Logistic regression, Bayesian theory, and support vector machine to simulate the impact of different types of data on real-time crash risk modeling. E conclusions of this paper can be used as a reference for the subsequent practice and research of highway traffic safety
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
