To study the influence of the driving environment of an undersea tunnel on driver EEG (electroencephalography) characteristics and driving safety, a real vehicle experiment was performed in the Qingdao Jiaozhou Bay Tunnel. The experimental data of the drivers’ real vehicle experiment were collected using an illuminance meter, EEG instrument, video recorder and other experimental equipment. The undersea tunnel is divided into different areas, and the distribution law of driving environment characteristics, EEG characteristics and vehicle speed characteristics is analyzed. The correlations between the driving environment characteristics, EEG characteristics and vehicle speed characteristics model the variables that pass the correlation test. The driving safety evaluation model of an undersea tunnel is established, and the driving safety in different areas of the undersea tunnel is evaluated. The results show that there are obvious differences in illumination, EEG power change rate, vehicle speed and other variables in different areas of the undersea tunnel. The driving environment characteristics are highly correlated with the β wave power change rate. The driving safety of different areas of the undersea tunnel from high to low is: upslope area, downslope area, exit area and entrance area. The study will provide a theoretical basis for the safe operation of the undersea tunnel.

The increase in the number of traffic accidents due to the increasing number of urban underpass tunnels necessitate a better investigation of drivers’ visual characteristics when entering the tunnels. A total of 20 drivers were gathered to perform a real vehicle experiment in an urban underpass tunnel. The saccade angle, saccade frequency, and fixation time were selected as the research indexes. The urban underpass tunnel entrance was divided into five sections, namely the external straight line section, the upper half of the ramp, the lower half of the ramp, the shading shed section, and the entrance inner section. The results showed that the saccade angle and frequency of the ramp were significantly smaller than that of the external straight line and the tunnel interior, and the saccade range in front of the entrance was more concentrated. The changes in fixation time and the difference range of 15th-85th fixation time threshold in each section were analyzed. The fixation time of all sections was distributed within the range of 149.476 to 475.414 ms. The driver’s fixation was more and more concentrated when the sidewalls were higher and closer to the portal.

This study aims to investigate the impact of the urban undersea tunnel longitudinal slope on the visual characteristics of drivers. 20 drivers were enrolled to conduct the real vehicle test of the urban undersea tunnel. First, the data of average fixation time and visual lobe were collected by an eye tracker. The differential significance was tested using the one-way repeated measures analysis of variance (ANOVA). Then, the difference between the up-and-down slope (direction) factor and the longitudinal slope (percent) factor on the two indexes were analyzed using the two-way repeated measures ANOVA. Second, by constructing a Lorentz model, the impact of the longitudinal slope on the average fixation time and the visual lobe were analyzed. Besides, a three-dimensional model of the longitudinal slope, average fixation time, and visual lobe was quantified. The results showed that the average fixation time and visual lobe under different longitudinal slopes markedly differed when driving on the uphill and downhill sections. The average fixation time and visual lobe under two factors were markedly different. Moreover, with an increase in the longitudinal slope, the average fixation time exhibited a trend of increasing first then decreasing; the visual lobe exhibited a trend of decreasing first and then increasing. The average fixation time reached the minimum and maximum value when the slope was 2.15% and 4.0%, whereas the visual lobe reached the maximum and minimum value when the slope was 2.88% and 4.0%. Overall, the longitudinal slope exerted a great impact on the visual load of the driver.

Traffic accidents at urban intersections may lead to severe traffic congestion, necessitating effective detection and timely intervention. To achieve real-time traffic accident monitoring at intersections more effectively, this paper proposes an urban road intersection accident detection method based on Generative Adversarial Networks (GANs), Extreme Gradient Boosting (XGBoost), and the SHAP interpretability framework. Data extraction and processing methods are described, and a brief analysis of accident impact features is provided. To address the issue of data imbalance, GAN is used to generate synthetic accident samples. The XGBoost model is then trained on the balanced dataset, and its accident detection performance is validated. In addition, SHAP is employed to interpret the results and analyze the importance of individual features. The results indicate that the accident samples generated by GAN not only retain the characteristics of real data but also enhance sample diversity, improving the AUC value of the XGBoost model by 7.1% to reach 0.844. Compared with the benchmark models mentioned in the study, the AUC value shows an average improvement of 7%. Additionally, the SHAP model confirms that the time–vehicle ratio and average speed are key factors influencing the model’s detection results. These findings provide a reliable method for urban road intersection accident detection, and accurate accident location detection can assist urban planners in formulating comprehensive emergency management strategies for intersections, ensuring the sustainable operation of traffic flow.

Urbanization has intensified traffic loads, posing significant challenges to the efficiency and stability of urban road networks. Overloaded nodes risk congestion, thus making accurate intersection importance classification essential for resource optimization. This study proposes a hybrid clustering method that combines Self-Organizing Maps (SOMs), K-Means, and the Gaussian Mixture Model (GMM), which is supported by the Traffic Flow–Network Topology–Social Economy (TNS) evaluation framework. This framework integrates three dimensions—traffic flow, road network topology, and socio-economic features—capturing six key indicators: intersection saturation, traffic flow balance, mileage coverage, capacity, betweenness efficiency, and node activity. The SOMs method determines the optimal k value and centroids for K-Means, while GMM validates the cluster membership probabilities. The proposed model achieved a silhouette coefficient of 0.737, a Davies–Bouldin index of 1.003, and a Calinski–Harabasz index of 57.688, with the silhouette coefficient improving by 78.1% over SOMs alone, 65.2% over K-Means, and 11.5% over SOM-K-Means, thus demonstrating high robustness. The intersection importance ranking was conducted using the Mahalanobis distance method, and it was validated on 40 intersections within the road network of Zibo City. By comparing the importance rankings across static, off-peak, morning peak, and evening peak periods, a dynamic ranking approach is proposed. This method provides a robust basis for optimizing resource allocation and traffic management at urban intersections.