The emergence of connected and autonomous vehicles (CAV) is of great significance to the development of transportation systems. This paper proposes a multiple-factors aware car-following (MACF) model for CAVs with the consideration of multiple factors including vehicle co-optimization velocity, velocity difference of multiple PVs, and space headway of multiple PVs. The Next Generation Simulation (NGSIM) dataset and the genetic algorithm are used to calibrate the parameters of the model. The stability of the MACF model is first theoretically proved and then empirically verified via numerical simulation experiments. In addition, the VISSIM software is partially redeveloped based on the MACF model to analyze mixed traffic flows consisting of human-driven vehicles and CAVs. Results show that the integration of CAVs based on the MACF model effectively improves the average velocity and throughput of the system.

The densely distributed, scale-varying objects in unmanned aerial vehicle (UAV) images, together with their dynamic, diverse, and unconstrained backgrounds, make conventional detection methods prone to missed detections, false alarms, and localization biases. To improve UAV vision tasks, we propose AD-YOLO, a unified method tailored for small object detection in traffic-dense settings. First, a module combining an adaptive rotation convolution unit and grouped directional attention with mixed-kernel features is introduced to enhance the model’s orientation invariance and multi-scale discrimination. Then, a dual-path collaborative feature pyramid network is proposed to jointly refine the model’s semantic and spatial details via a multi-directional context aggregation path and a hierarchical semantic progressive fusion path. Last, a hierarchically dense reparameterized large-kernel module is designed to produce broader receptive fields with reduced computational complexity. Extensive experiments on the VisDrone2019 and UAVDT datasets demonstrate that AD-YOLO outperforms state-of-the-art methods in detection accuracy while maintaining favorable computational efficiency.

As a kind of brand-new traffic facility, the role of underground expressways in a traffic system is usually illustrated by sharing rates which are limited by travel distance, time, speed, cost and other factors. A Logit model is often used to obtain a sharing rate, however, the fatal disadvantage of Logit usually causes imprecision. This paper proposes using a Mixed Logit model to obtain a sharing rate considering the above four factors, selects utility data from different travel scenes in the network of an underground expressway and conventional road and uses a simulation algorithm of statistical analysis software SAS to calibrate utility function parameters. Finally, this study determines the sharing rate of an underground expressway form. The research shows that the sharing rate based on a Mixed Logit model is more reasonable and advantaged for evaluating the effects of an underground expressway.

Accident data collected from 2014 to 2019 for the Dongguan section of the Guangshen Yanjiang Expressway in China were utilized to investigate the key factors influencing the severity of road property damage in highway traffic accidents. The spatial correlation among adjacent accidents was addressed using a spatial generalized ordered Probit model, which employed varying association distance thresholds. An XGBoost machine learning algorithm was developed to estimate the model parameters, and the SHAP (SHapley Additive exPlanations) method was employed to elucidate the model outputs. The results show that significant spatial correlations are present within the accident data. The spatial generalized ordered Probit model demonstrated superior performance compared to the conventional generalized ordered Probit model, with the model based on a 200 m association distance threshold yielding the best results. The SHAP method significantly enhanced the interpretability of the XGBoost machine learning model. Parameter estimation revealed that variables such as single-vehicle accidents, passenger cars, lorries, heavy tractors, nighttime occurrences, early morning periods, cloudy conditions, rainy conditions, and bridge locations were significantly associated with the severity of road property damage resulting from traffic accidents.

Pick/drop stand road (PDSR) is necessary for specific functions, but it very easily generates congestion and thus results in a less safe and comfortable ride for passengers. The Internet of Vehicles has offered new ideas and technical means to solve the problem. First, this study divides the PDSR into different functional zones and constructs the coordinated control environment of connected vehicles (CVs). A car-following coordinated control strategy (CF-CCS) and a lane-changing coordinated control strategy (LC-CCS) are proposed based on a mixed platoon of CVs and human-driven vehicles. For CF-CCS, this study offers a model for car-following coordinated platoon control based on terminal state prediction of the leading vehicle, with platoon safety and passengers’ comfort as optimization objectives, and solves it using a modified particle swarm algorithm. For LC-CCS, a model for lane-changing coordinated control of platoon is proposed, and a numerical approach is utilized to solve it. It is based on finding the best area for lane-changing location while considering the safety of lane changing and the efficiency of vehicle traffic on the PDSR. Last, VISSIM and MATLAB simulations are conducted to prove that the proposed two strategies can effectively ease congestion, significantly increase the speed of traffic flow on PDSR, and provide passengers with safer and more comfortable rides.