Different operating environments under the action of crosswind lead to an abrupt change in the aerodynamic characteristics of high-speed trains (HSTs), which seriously affect the train operation safety and passenger comfort. Considering the compressibility and unsteady characteristics of flow field, a 3D numerical model including tunnel, HST and crosswind was established, and the SST k-w model was adopted to solve the problem. The accuracy of the numerical simulation was verified by comparing with the dynamic model test. It further analyzed the influence of cross wind on the flow field and surface pressure distribution around the train, and obtained the aerodynamic load change law of the train under the action of cross wind. The results show that the flow field distribution around the train is significantly affected by the crosswind. The flow field shifts to the leeward side of the train outside the tunnel, forming a longitudinal vortex starting from the tunnel entrance, while the vortex structures on the leeward side of the train disappear and form a vertical vortex at the extension entrance in the space on the windward side. Furthermore, the vortex structures in the tunnel disappear as the train enters. Before the train enters the tunnel, the aerodynamic pressure on the windward surfaces of the train is mainly positive, and the aerodynamic pressure on the leeward surfaces is mainly negative.The surface pressure of the train changes most obviously when the train enters the tunnel, and the fluctuation degree of the aerodynamic pressure decreases obviously with the train entering the tunnel. The variation of aerodynamic load is closely related to the wind environment. The side force and lift amplitude of the rear vehicle (RV) are the largest when there is non-crosswind, and the side force and lift amplitude of the head vehicle (HV) are the largest when there is crosswind. In addition, the aerodynamic performance is closely related to the marshaling position. The variation amplitudes of the side force of the HV are 4.8 and 15.4 times of that of the RV, respectively. The variation amplitudes of the lift of HV are 1.1 and 1.2 times of that of RV, respectively. And the risk of traveling safety accidents of the HV is the highest. The research results can provide a reference for the safety evaluation of HST and route selection of high-speed railway tunnels.

The sensor of unmanned vehicles replaces human eyes to perceive the information of road space is an important prerequisite for the safe operation of unmanned vehicles. Therefore, based on the domestic and international research literature research as well as the analysis of relevant software and hardware technologies, this paper analyzed the limitations of the current perception technology of unmanned vehicles, including the recognition range and characteristics of sensors such as LiDAR, camera and millimeter-wave radar. And it collected the real information of the road by using the LiDAR and the combination of navigation system, and further constructed the three-dimensional perception space of the unmanned vehicle with the collected point cloud data, the localization information, and the synchronous positioning and modeling algorithms, which realize the three-dimensional digital model construction of the road. At the same time, the mathematical expression model construction of parameters such as ranging ability, horizontal field of view angle and vertical field of view angle was carried out in the 3D point cloud map, and the spatial coordinate transformation method was utilized to separate the 3D point cloud data within the recognition range of the sensors and convert them to a unified coordinate system. Finally, the Delaunay triangulation method was used to construct a 3D model that can reflect the characteristics of perceptual space, so as to realize the perceptability of 3D perceptual space. In order to verify the practicality and accuracy of this method, this paper tested the algorithm using data collected in the field. The test results show that the method proposed in this paper has good robustness and it can work stably in complex road environments and accurately assess the perceptibility of unmanned vehicles. This research result not only provides a scientific basis for the road design and safety assessment of unmanned vehicles, but also provides strong technical support for the further development and application of unmanned technology.

In order to improve traffic operation of freeway bridge sections in fog, this paper proposed a human factor-oriented operation risk active warning prevention and control strategy for bridges. This strategy integrates traditional traffic safety facilities with intelligent technologies such as risk warning, dynamic speed limits, and visual guidance. Through changes in prevention and control levels under different risk levels, the facility resources were fully utilized to achieve the goal of multi-level and differentiated prevention and control from static notification to dynamic guiding. However, it is not yet clear whether this process will improve the driving behavior and the safety and efficiency of traffic operation, so it is difficult to support the effective design and application of the prevention and control strategy. Therefore, E’dong Yangtze River Bridge was taken as an example to quantitatively evaluate the improvement effectiveness of the prevention and control strategy through driving simulation test and to verify its validity. The optimization and implementation is based on the fact that human factor needs are subjected to the prevention and control strategy, then it has a better effect in guiding the practical application. The improvement effectiveness evaluation of prevention and control strategy was realized through the improvement rate of each indicator from three aspects (psychological perception, safety effectiveness and traffic efficiency). The results show that the prevention and control strategy is helpful to improve driver’s confidence and alleviate tension; it can significantly improve the driver’s behavioral performance and traffic operation state and enhance the traffic efficiency and the safety of horizontal and longitudinal operation of vehicles; it has the strongest ability to improve safety effectiveness, especially at the longitudinal level; the higher brightness line-of-sight induction strategy is more conducive to improving the safety level and significantly increasing the traffic capacity. Compared with the low brightness condition, it may have a negative impact on the driver’s psychological feelings, but it is not obvious. When the visibility is 100 m on bridge, it is recommended that the flashing frequency of visual guidance technology is 0.5 Hz and the brightness is 3 500 cd/m². The results provide the necessary theoretical basis for the practical application of prevention and control strategy on foggy bridges. The proposed solution and the analytical idea of its improvement effectiveness provide a reference for targeted prevention and management of risky sections.

The active management and control strategy in the weaving section of expressways plays a crucial role in enhancing operational efficiency and improving driving safety. Self-luminous road marking is a new type of road traffic safety active control facility. Because of its intelligent control, virtual and real transformation, emergency warning and other functional characteristics, it has attracted more and more attention and gradually promoted and applied. In order to explore the effect of the new road marking in the expressway weaving area, based on the driving simulation test, this paper analyzed the influence characteristics of self-luminous road marking on driving behavior from the perspective of drivers and evaluated the effect of self-luminous road marking. Two simulation scenarios were designed: one with conventional road markings and the other with self-luminous road markings under nighttime conditions. Driving behavior data were collected, and six indicators were selected to construct an evaluation system, focusing on driving safety, comfort, and driver control performance. These indicators include standard deviation of speed, standard deviation of lateral position, root mean square of longitudinal acceleration, standard deviation of acceleration, mean speed, and throttle effectiveness. Further analysis was conducted to examine the significance of differences and effect size levels of various indicators under different road marking scenarios. Finally, an extension matter-element model was applied to construct a comprehensive evaluation model to compare the effects of the two road marking schemes. The results indicate that self-luminous road markings significantly influence driving behavior characteristics in weaving sections. Under the self-luminous marking condition, drivers’ lateral position perception and control levels is improved, speed regulation ability is increased, and vehicle movement becomes more stable. The comprehensive evaluation based on the extension matter-element model shows that self-luminous road markings are more effective than conventional road markings in enhancing driving safety in expressway weaving sections.

The connection between urban rail transit and bus transit is the key to meet the various urban travel demand and to promote the development of urban public transportation system. Existing studies lack the consideration of some important micro-indicators, such as the ratio of waiting for multiple buses during peak hours and the level of congestion inside the bus, when constructing the optimization model. Additionally, there is a lack of consideration for the stochastic and heterogeneous requirements in route operation, which results in poor performance in practical applications. To address these issues, this study firstly established an optimization model based on the service process of the bus transit, with the objective of minimizing the travel cost of passengers and the cost of enterprises. The model considers the influencing factors such as operating speed, vehicle type, departure frequency, route fare, vehicle crowdedness, and route line type and it is solved by the non-dominated sorting genetic algorithm (NSGA-Ⅱ), in which the genetic operation part is improved. Furthermore, a microscopic simulation algorithm was designed to evaluate the solution in order to improve the accuracy of the model solution. Accordingly, a Kriging surrogate model was used to assist the calculation to improve the solution efficiency of the algorithm. Finally, taking the connection between metro and bus system in Shenzhen city as an example, the proposed algorithm was validated with the IC card data collected in metro and bus system. The sensitivity analysis was conducted for the factors of route fare, operating speed, operating mode and passenger volume, and the operating improvement was proposed based on the analysis results. The results demonstrate that the algorithm produces superior route solutions compared to the conventional NSGA-Ⅱ, with the same solving time. There is a 35.49% reduction in total cost and a notable 26.94% increase in the iteration speed. The optimization method for connecting between metro and bus transit proposed in this study has practical significance in improving connecting efficiency and operational level.

The traffic modes in metropolis is interwoven to form an interconnected passenger flow network, and the spatio-temporal relationship between cross-transportation modes is complicated, which requires the joint prediction of passenger flow to analyze the overall travel law. For the cross-transportation passenger flow network, the supergraph correlation matrix of cross-transportation modes is introduced to describe the correlation between the passenger flow supergraph network of bus and subway, and a joint prediction model based on dual-mode spatial-temporal supergraph convolution network (BSTHCN) is proposed. Specifically, the model consists of three parts: an input module, a spatio-temporal convolution module (including temporal and spatial convolutions), and an output module, which can simultaneously capture the passenger flow network characteristics of both buses’ and metros’ stations and routes, as well as the transfer passenger flow characteristics between the two different passenger flow networks. The proposed model can identify and extract important information features, and perform feature aggregation and allocation. The experimental results show that the proposed model has better prediction accuracy compared to classical prediction models. The proposed model reduces the mean absolute error (MAE) by 8.93% and 8.10% on the bus and metro datasets, respectively, while the RMSE decreased by 10.64% and 7.47%. Moreover, the parameter volume and model runtime of proposed model are within a reasonable range. Compared to S-TGCN and DCRNN, proposed model achieves more accurate predictions with only a 4.82% increase in runtime. On the whole, proposed model demonstrates strong competitiveness. The ablation experimental results further demonstrate that after incorporating hypergraphs and considering multi-modal transportation correlations, the proposed model can better reflect both local and global characteristics in passenger flow networks, thereby improving the accuracy of passenger flow prediction.