Vehicle-to-Everything (V2X) communication is expected to accomplish a long-standing goal of the Connected and Autonomous Vehicle (CAV) community to bring connected vehicles to roads on a large scale. A major challenge, and perhaps the biggest hurdle on the path towards this goal, is the scalability issues associated with it, especially when vehicular safety is concerned. As a major stakeholder, Cellular V2X (C-V2X) community, which is based on the 3rd Generation Partnership Project (3GPP), has long been trying to research on whether vehicular networks are able to support the safety-critical applications in high-density vehicular scenarios. This paper attempts to answer this question by first presenting an overview on the scalability challenges faced by 3GPP Release 14 Long Term Evolution C-V2X (LTE-V2X) using the PC5 sidelink interface for low and heavy-density traffic scenarios. Next, it demonstrates a series of solutions that address network congestion, packet losses, and other scalability issues associated with LTE-V2X to enable this communication technology for commercial deployment. In addition, a brief survey is provided into 3GPP Release 16 5G New Radio V2X (NR-V2X) that utilizes the NR sidelink interface and works as an evolution of C-V2X towards better performance for V2X communications, including new enhanced V2X (eV2X) scenarios that possess ultra-low-latency and high-reliability requirements.

With the spread adoption of artificial intelligence, the great challenges confronted by the intelligent safety concern-safety of the intended functionality has become the biggest roadblock to the mass production of high-level automated vehicles, notably arising from perception algorithm deficiencies. This paper focuses a cut-in scenario, dividing this scenario into low-risk and high-risk segments predicated on the kinetic energy field, and the mental activities of passengers on prefrontal cortex, are analyzed within these delineated segments. Two experiments are then conducted, leveraging driving simulators and real-world vehicles, respectively. Experiment results indicate that high risk may result in the passengers’ mental activity on prefrontal cortex change. This revelation posits a potential avenue for augmenting the intended functionality of automated vehicle by using passengers’ physiological state.

Testing and evaluation plays a critical role in the research and development (R&D) of autonomous vehicles (AVs). Due to the black-box property of AVs and the restraint of test resources, how to quickly test and evaluate AVs’ safety remains a major challenge. To address this problem, a novel search-based ADOE testing and evaluation method is proposed, which improves the efficiency of testing and evaluation through a two-stage acceleration. In Accelerated Testing stage, the proposed ADOE-based testing method (DUSGAT) is used to accelerate testing, which adopts dual surrogates (result surrogate M_R and behavior surrogate M_B). This method indicates not only the previous searching results but also the behavior pattern of the system under test (SUT) during previous tests, to accelerate the whole process of unveiling all critical scenarios for the SUT. In the Accelerated Evaluation stage, the trained M_B is used to quickly predict and evaluate the safety performance of SUT in the logical scenario. Experimental results show that DUSGAT has the best search performance compared with baseline methods, and the F2-score of DUSGAT is higher than TuRBO. In the 2-dimensional car-following scenario, the relative error between the output SUT’s hazard ratio of our method and that of grid search is only 0.68%. In the 3-dimensional cut-in scenario, the relative error is only 1.32%. What's more, compared with grid search, this study’s method is 3.217?×?faster in the car-following scenario, and 22.116?×?faster in the cut-in scenario. Therefore, the method can accurately test and evaluate the safety performance of SUT and has the potential to be used in high-dimensional scenarios.

Merging into the highway from the on-ramp is an essential scenario for automated driving. The decision-making in this scenario needs to balance safety and efficiency to optimize a long-term objective, which is challenging due to the dynamic, stochastic, and adversarial characteristics. The existing learning-based methods struggle to meet the safety requirements. This paper proposes a reinforcement-learning-based decision-making method under a framework of offline training and online correction, called the Shielded Distributional Soft Actor-critic (Shielded DSAC). The Shielded DSAC adopts the policy evaluation with safety considerations in offline training, and a safety shield parameterized with the barrier function in online correction. These two measures support each other in achieving better safety without sacrificing efficiency performance. The study verified the Shielded DSAC in a simulated on-ramp merge scenario. The results indicate that the Shielded DSAC has the best safety performance compared to baseline algorithms and achieves efficient driving simultaneously.

In order to improve the performance of object detection algorithm in dynamic traffic scenarios, a moving traffic object detection method based on Bayesian theory fusion is proposed. To obtain initial object detection results, adaptive coordinate attention YOLO (ACA-YOLO) network with high accuracy and multi-scale optical flow (MSOF) method with high sensitivity to dynamic object are applied, respectively. To enhance the detection performance of YOLOv5 network, adaptive coordinate attention (ACA) mechanism is applied to obtain more accurate location and identification of interest objects. To address the issue of constant loss values when the inclusion relationship between truth boxes and prediction boxes occurs, the loss function now utilizes efficient-IoU instead of generalized-IoU. Fusion weights are obtained by calculating the posterior probabilities of ACA-YOLO network and MSOF method separately using Bayesian formula after object detection regions matching based on intersection over union (IoU). The results of fusion detection are based on the posteriori probabilities. The proposed detection method was tested on the KITTI dataset and self-built continuous moving traffic objects dataset which consists of real continuous dynamic traffic scenarios. The experimental results indicate that the proposed method for detecting moving traffic object based on Bayesian theory fusion has outstanding performance. The mean average precision, Recall and Precision of the proposed method on KITTI dataset reach 0.954, 0.912, 0.944, which are 5.5%, 9.9% and 1.9% higher than that of traditional YOLOv5 network.

Vehicle taillight signals contain a wealth of semantic information crucial for inferring the leading vehicle's driving intentions. In this paper, to enhance the recognition accuracy, optimize the hardware requirements for deployment, and reduce the model reasoning time, a lightweight method for taillight signal recognition is proposed. This method involves three stages: detection, tracking, and recognition. Firstly, the lightweight MCA-YOLOv5 network is designed for the detection of vehicle rears. Subsequently, the detection results are tracked via the Bytetrack algorithm, resulting in tracking sequences. Finally, the TSA-X3d network aims to effectively obtain spatio-temporal information from the tracking sequences and recognize the taillight signals. The experimental results indicate that the MCA-YOLOv5s network significantly outperforms its precursor—the original YOLOv5s model—in efficiency and size. Specifically, the model size, parameter count, and computational demand of the proposed MCA-YOLOv5s network are respectively reduced to 33.33%, 31.43%, and 34.38% of those of the original YOLOv5s, yet it maintains comparable average precision. Furthermore, when compared with other typical taillight signal recognition algorithms, the TSA-X3d network not only has the fewest number of parameters, but also achieves the highest accuracy, reaching 95.39%. To mitigate deployment challenges, the study leverages TensorRT to markedly decrease the inference time of the MCA-YOLOv5s model to 1–3 ms. Additionally, it employs quantization techniques on the TSA-X3d model, slashing its inference time to 30% of the original and shrinking its model size to 26.65% of its initial footprint. Notably, the augmented algorithm's inference speed surpasses 25 frames per second, surpassing the threshold necessary for real-time taillight signal recognition, thus showcasing its potential for immediate practical application.

Accurate and efficient pedestrian trajectory prediction is one of the key capabilities for the safe operation of self-driving vehicles. Therefore, it is of great significance to study pedestrian trajectory prediction algorithms applicable to complex interaction scenarios. In this study, a spatial gridding-based multi-attention generative adversarial network (SGMA-GAN) is proposed, which is modeled with generative adversarial network as the main framework. Firstly, the map information is gridded to better represent the pedestrian state information in tensor form, improve the stability of the state space and network structure. Secondly, temporal and spatial attention mechanisms are introduced to account for the effects of historical trajectories and spatial interaction features. Finally, the model is evaluated with both Eidgen?ssische Technische Hochschule (ETH) and University of Cyprus (UCY) datasets. The results showed that as the prediction step size gradually increased, compared with the relatively new SGANv2, the mean average displacement error (ADE) and Final displacement error (FDE) of SGMA-GAN in five scenarios increased by 10.61% and 4.65%, respectively.

Advanced research in 5G-enabled vehicular technologies has made the Internet of Vehicles (IoV) a promising research area in intelligent transportation systems (ITS). The high vehicular mobility causes frequent network topology changes, leading to unreliable emergency alert messaging (EAM) in real-time IoV applications. Due to the high volume of data propagation in such networks with limited bandwidth resources, topological dynamics can cause data congestion and data loss, which is not tolerable for delay-sensitive emergency messages (EMs). Unmanned Ariel vehicles (UAVs) as dynamic infrastructures can provide customized EAM services for IoVs. This paper proposes UFC3, a UAV-aided Fog Computing Congestion Control strategy for EM dissemination in 5G IoV. This study provides a practical method for EAM by reducing burst EM traffic to guarantee reliable messaging communication between an abnormal vehicle (AV) and the fog server (FS). In UFC3, various nodes capable of Wave protocol were considered as basic vehicles or 5G-enabled nodes regarded as intelligent vehicles or UAVs, and the authors propose a traffic-aware forward/backward link EAM strategy to handle any EM and the corresponding response between the AV and FS. Next, this paper suggests a mathematical analysis for EM propagation speed (EMPS) and propose closed-form EMPS equations for various traffic-aware forward/backward link EAM scenarios. Using OMNET?+??+?along with Veins and INET frameworks, the authors simulate UFC3 and compare it with previously published works regarding communication overhead, throughput, packet delivery ratio, average delay, packet loss ratio, channel utilization, and energy consumption.

Capacity is a crucial metric for evaluating the degradation of lithium-ion batteries (LIBs), playing a vital role in their management and application throughout their lifespan. Various methods for capacity estimation have been developed, including the traditional Ampere-hour integral method, model-driven methods based on equivalent circuit models or electrochemical models, and data-driven methods based on features extracted from partial charging, discharging, or relaxing processes. Current research focuses on improving the accuracy, acquisition speed, and robustness of these capacity estimation methods. This study proposes a rapid and precise method for capacity estimation in LIBs, using electrochemical impedance spectroscopy (EIS) and the extreme gradient boosting machine learning framework. The proposed method concurrently considers the impacts of the state of charge (SOC) and temperature. The model demonstrates the ability to automatically compensate for variations in SOC and temperature, leveraging specific impedance features, provided the input EIS data's SOC and temperature range is encompassed within the training set. Two implementations of the method are presented. The first utilizing EIS features, while the second employs features derived from the distribution of relaxation times. The latter exhibits enhanced adaptability to small datasets. When applied to the complete dataset of this study, the proposed method achieves an R2 value exceeding 0.97 and a mean absolute percentage error below 0.8%.

For the design of the hybrid electric vehicles, the strong coupling between plant parameters and controller parameters turns the problem into a multi-layered challenge. If handled sequentially, it is defined as sub-optimal. In order to obtain the optimal design of the system, it is necessary to integrate the physical system and its controller. Taking component parameters and energy management strategy as research objects, this paper elaborates an integrated optimization approach for a series–parallel hybrid electric vehicle. Firstly, a rule-based control strategy that can be applied online is designed according to various driving modes of the hybrid electric vehicle. Then, considering the coupling between component parameters and control strategies, a dual-layer optimization framework with genetic algorithm and double dynamic programming is proposed to optimize fuel economy and battery life. Among them, parameters of component size and control for the upper-layer of the framework are selected as preparative optimization parameters. In order to get rid of the influences of energy management strategies and obtain the optimal upper-layer parameters, the lower-layer of the framework adopts the global optimization algorithm to calculate the optimal energy distribution ratio for each driving mode. The results indicate that, while ensuring the good working condition of the battery, the fuel economy has improved by 7.79% under the selected driving cycle after optimization. The optimized upper-layer parameters combined with the proposed control rules can be applied online.

The electric wheel has an advantage of independently, accurately, and promptly controlling torque in response. However, current distributed drive steering control strategies fail to fully leverage this capability. To fill this gap, this paper proposes a composite electronic moment power steering (EMPS) control strategy for multi-axle distributed drive vehicles, based on dynamic modeling and analysis. The proposed control strategy integrates EMPS with direct yaw moment control (DYC), enhancing steering flexibility and fault tolerance in the steering system at low speeds, while also ensuring vehicle stability at high velocities. It adopts a hierarchical control architecture, wherein the upper controller utilizes a nonlinear state observer for the joint estimation of multi-objective parameters, and the lower controller is responsible for the accurate tracking of the steering angle by EMPS, the yaw rate tracking by DYC, and the assignment of weights to the composite controller. By establishing and analyzing a detailed vehicle model and an electric drive steering axle dynamics model, a multi-dimensional feasible domain for EMPS is proposed, ensuring the safety and smoothness of steering maneuvers. The co-simulation of MATLAB/Simulink and TruckSim are conducted to verify the effectiveness of the composite control strategy. The EMPS controller is proved to have robust steering angle tracking performance, with the β——β* trajectory consistently converging within the stability zone, maintaining a sufficient margin for tire longitudinal force under the composite steering control. Additionally, real vehicle testing confirms the effectiveness of EMPS and the redundancy tolerance of the steering functionality in distributed drive multi-axle vehicles equipped with EMPS.

Gearshift control poses significant influence on smoothness and dynamics of vehicles, especially for vehicles equipped with dual clutch transmissions (DCT). However, classical strategies are difficult to account for complex transient variations of the clutch friction coefficient during gear shifting. To cope with this challenge, this study proposes an adaptive gearshift control algorithm based on physical and data-driven modeling for regulation of the clutch friction coefficient. A closed-loop strategy is proposed for the speed control in the gearshift process of DCT vehicles. The reference speeds of clutch and engine are determined via online invocation of the pseudo-spectrum method. Meanwhile, an adaptive sliding mode controller based on the fused physical and data-driven hybrid model is designed to control the clutch oil pressure in the torque phase, thereby enabling precise tracking of the reference speed of the clutch. Then, an efficient engine controller is designed based on sliding mode control to regulate the throttle opening for following the engine reference speed in the torque phase. The simulation and experimental results manifest that the proposed adaptive control method maintains satisfactory shift quality with high adaptive capability and strong robustness under time-varying throttle pedal opening.