Dynamic short-term crash analysis and prediction at toll plazas for proactive safety management.

Toll plazas are commonly recognized as bottlenecks on toll roads, where vehicles are prone to crashes. However, there has been a lack of research analyzing and predicting dynamic short-term crash risk specifically at toll plazas. This study utilizes traffic, geometric, and weather data to analyze and predict dynamic short-term collision occurrence probability at mainline toll plazas. A random-effects logit regression model is employed to identify crash precursors and assess their impacts on the probability of crash occurrence at toll plazas. Meanwhile, a Long Short-Term Memory Convolutional Neural Network (LSTM-CNN) network is applied for crash prediction. The results of random-effects logit regression model indicate that the flow standard deviation of downstream, upstream occupancy, speed difference and occupancy difference between upstream and downstream positively influence the probability of crash occurrence. Conversely, an increase in the proportion of ETC lanes negatively impacts the probability of crash occurrence. Additionally, there appears a higher likelihood of crashes occurring during summer at toll plaza area. Furthermore, to address the issue of data imbalance, Synthetic Minority Oversampling Techniques (SMOTE) and class weight methods were employed. Stacked Sparse AutoEncoder-Long Short-Term Memory (SSAE-LSTM) and CatBoost were developed and their performance was compared with the proposed model. The results demonstrated that the LSTM-CNN model outperformed the other models in terms of the Area Under the Curve (AUC) values and the true positive rate. The findings of this study can assist engineers in selecting suitable traffic control strategies to improve traffic safety in toll plaza areas. Moreover, the developed collision prediction model can be incorporated into a real-time safety management system to proactively prevent traffic crash.

Safety improvements by intelligent connected vehicle technologies: A meta-analysis considering market penetration rates.

As the era of intelligent connected vehicles (ICVs) is approaching, a number of studies have investigated the potential benefits of ICVs, including the safety effects. Although previous studies agree that ICVs would significantly improve traffic safety, its quantified safety effects at different stages are still being debated. This study aims to estimate the ICVs' safety effects by market penetration rate (MPR) adopting a meta-analysis approach. The results from the meta-analysis indicate that the number of conflicts is exponentially reduced as the MPR goes up. For example, compared to the environment without ICVs, 4.2% and 17.4% of conflicts would decrease at the MPR of 10% and 50%, respectively. The effects are more obvious at higher MPR-43.4% of conflicts are expected to decrease at the MPR of 90%. From the case study in the United States based on the meta-analysis, it is expected that the MPR would reach 17-20% in the near future (2025) and 40-48% in 2035. The anticipated reduction in the number of fatal crashes would be 5% and 13%, in 2025 and 2035, respectively. The findings from this study will be useful for both public and private sectors to establish strategic plans to promote ICVs and identify their benefits at different MPRs.

The extended theory of planned behavior considering heterogeneity under a connected vehicle environment: A case of uncontrolled non-signalized intersections.

With the emergence of connected vehicle (CV) technology, there is a doubt whether CVs can improve driver intentions and behaviors, and thus protect them from accidents with the provision of real-time information. In order to understand the possible impacts of the real-time information provided by CV technology on drivers, this paper aims to develop a model which considers the heterogeneity between drivers with the aid of the extended theory of planned behavior. At the uncontrolled non-signalized intersections, a stated preference (SP) questionnaire survey was conducted to build the dataset consisting of 1001 drivers. Based on the collected dataset, the proposed model examines the relationships between subjective norms, attitudes, risk perceptions, perceived behavioral control and driving intentions, and studies how such driving intentions are simultaneously related to driver characteristics and experiences in the CV environment. Furthermore, driver groups which are homogenous with respect to personality traits are formed, and then are employed to analyze the heterogeneity in responses to driving intentions. Four key findings are obtained when analyzing driver responses to the real-time information provided by CV technology: 1) the proposed H-ETPB model is verified with a good fitness measure; 2) irrespective to driver personality traits, attitudes and perceived behavioral control have a direct and indirect association with driving intentions to accelerate; 3) driving intentions of high-neurotic drivers to accelerate are significantly related to subjective norms, while that of low-neurotic drivers are not; 4) elder high-neurotic drivers, and low-neurotic drivers who have unstable salaries or ever joined in online car hailing service have a strong intention in accelerating. The findings of this study could provide the theoretical framework to optimize traffic performance and information design, as well as provide in-vehicle personalized information service in the CV and CAV environments and assist traffic authorities to design the most acceptable traffic rules for different drivers at an uncontrolled non-signalized intersection.