Modeling the influence of connected vehicles on driving behaviors and safety outcomes in highway crash scenarios across varied weather conditions: A multigroup structural equation modeling analysis using a driving simulator experiment
Abdalziz Alruwaili,Kun Xie
DOI: https://doi.org/10.1016/j.aap.2024.107514
IF: 5.9
2024-02-24
Accident Analysis & Prevention
Abstract:Equipped with advanced sensors and capable of relaying safety messages to drivers, connected vehicles (CVs) hold the potential to reduce crashes. The goal of this study is to assess the impacts of CV technologies on driving behaviors and safety outcomes in highway crash scenarios under diverse weather conditions, including clear and foggy weather. A driving simulator experiment was conducted and the multigroup structural equation modeling (SEM) was employed to explore the complex interrelationships between the propensity of traffic conflicts, utilization of CV alerts, weather, psychological factors, driving behaviors, and other relevant variables for two different crash locations, namely a straight section and a horizontal curve. Two latent psychological factors including aggressiveness and unawareness were constructed from driving behavior as vehicles passed by crash scenes such as brake, throttle, steering angle, lane offset, and yaw. The SEM can measure latent psychological factors and model interrelationships concurrently through a single statistical estimation procedure. Results of the multigroup SEM showed that CV alerts could significantly reduce the unawareness on a horizontal curve and thus lower the propensity of traffic conflicts. Additionally, the overall effect of foggy weather on conflicts was found to be positive on a horizontal curve, despite the potential benefit of improving situational awareness. In contrast, the single group SEM failed to reveal any significant interrelationships in its structural model by pooling data from both crash locations. The obtained insights can guide the development of driving assistance systems, highlighting the necessity of customization considering weather conditions and location-specific factors.
public, environmental & occupational health,transportation,ergonomics,social sciences, interdisciplinary