Abstract:Identifying critical safety management drivers with high driver-level risks is essential for traffic safety improvement. Previous studies commonly evaluated driver-level risks based upon aggregated statistical characteristics (e.g., driving exposure and driving behavior), which were obtained from long-period driving monitoring data. However, given the great advancements of the connected vehicle and in-vehicle data instrumentation technologies, there has been a notable increase in the collection of short-period driving data, which has emerged as a prominent data source for analysis. In this data environment, traditionally employed aggregated behavior characteristics are unstable due to the time-varying feature of driving behavior coupled with insufficient data sampling periods. Thus, traditional modeling methods based upon aggregated statistical characteristics are no longer feasible. Instead of utilizing such unreliable statistical information to represent driver-level risks, this study employed temporal variation characteristics of driving behavior to identify critical safety management drivers in the short-period driving data environment. Specifically, the relationships between driving behavior temporal variation characteristics and individual crash occurrence probability were developed. To eliminate the impacts of drivers' driving behavior heterogeneity on model performance, "traffic entropy" index that could quantify the abnormal degrees of driving behavior was proposed. Deep learning models including convolutional neural network (CNN) and long short-term memory (LSTM) were employed to conduct the temporal variation feature mining. Empirical analyses were conducted using data obtained from online ride-hailing services. Experiment results showed that temporal variation characteristics based models outperformed traditional aggregated statistical characteristics based models. The area under the curve (AUC) index was improved by 4.1%. And the proposed traffic entropy index further enhanced the model performance by 5.3%. The best model achieved an AUC of 0.754, comparable to existing approaches utilizing long-period driving data. Finally, applications of the proposed method in driver management program development and its further investigations have been discussed.

Estimation of Driver's Driving State Based on Cloud Model

Vision‐based Inter‐vehicle Distance Estimation for Driver Alarm System

Research and Implementation of Networked Cloud-Controlled Autonomous Driving System

Abnormal Driving Detection Based on Normalized Driving Behavior

Prediction Analysis of Traffic Flow Based on Cloud Model

Driving state evaluation of intelligent connected vehicles based on centralized multi-source vehicle road collaborative information fusion

Aware of Scene Vehicles—Probabilistic Modeling of Car-Following Behaviors in Real-World Traffic

Vehicle Motion State Prediction Method Integrating Point Cloud Time Series Multiview Features and Multitarget Interactive Information

Driving with knowledge from the physical world.

Cloud-model Based Traffic State Identification on Urban Expressway

A detection model for driver's unsafe states based on real-time face-vision

A Context-Aware Evaluation Method of Driving Behavior.

A hierarchical fuzzy decision model for driver's unsafe states monitoring.

Scene-aware Driver State Understanding in Car-Following Behaviors

Unified Quantitative Method of Driving Risk by Comprehensive Considering Driver-Vehicle-Road Factors

Simulated Analysis of Modeling of Driving Behavior Characteristics Based on Satellite Positioning Data.

Driver State Monitoring Technology for Conditionally Automated Vehicles: Review and Future Prospects

A Driving Risk Surrogate and Its Application in Car-Following Scenario at Expressway

Driving Safety Prediction and Safe Route Mapping Using In-vehicle and Roadside Data

Critical safety management driver identification based upon temporal variation characteristics of driving behavior

Vehicular Networking-Enabled Vehicle State Prediction via Two-Level Quantized Adaptive Kalman Filtering