Crack detection using deep learning (DL) is important for reducing infrastructure operation risks, saving operation and maintenance costs, and promoting the intelligent transformation of the civil engineering industry. In current practice, algorithms, datasets, and evaluation metrics are the key components of DL-based crack detection. Hence, in this article, four aspects of current research are reviewed systematically. First, the development of the DL method is reviewed, and the applications of deep convolution neural networks in the field of computer vision and their significant advantages over conventional algorithms in image data processing are introduced. Second, three popular DL algorithms for crack detection are described in detail. Third, the available crack image datasets and current evaluation metrics are reviewed. Finally, recent research outcomes in this field are summarized, and future research needs are discussed. The comprehensive analyses show that, based on the backbone of convolutional neural networks, DL algorithms have been widely used to locate and classify cracks on the surface of civil infrastructure with good accuracy, although obtaining quantitative information about cracks still requires auxiliary extraction using traditional image-processing technology. Because of the high cost and technical skills required for pixel-level annotation, there is a lack of large-scale crack semantic segmentation datasets, resulting in crack detection models with poor robustness. Moreover, most researchers created their own datasets for model training and used different metrics to evaluate their model performance, highlighting the need to establish a benchmark dataset for model training and use a set of popular indices to compare the performances of different models. Crack detection robots have been developed for different types of infrastructure, and it is the development trend to improve multiscene adaptability and reduce the application cost of crack detection robots.

Road information is the key to the decision-making of digital traffic infrastructure and maintenance behavior. The application of advanced detection equipment is the way to obtain road information. The high-efficiency, and non-destructive detection characteristics of 3D Ground Penetrating Radar (3D GPR) can support the acquisition of road structural defects data. This paper reviews the typical types of road structural defects, detection methods, etc. The major principles of 3D GPR technology, and its application on road engineering are introduced afterwards. Subsequently, this paper concludes the application and development of artificial intelligence technology in GPR image recognition technology according to the structural defect characteristics and recognition methods. According to the development of traffic infrastructure, this paper looks forward to the digital twin technology based on 3D GPR data, mainly introduces the modeling and model simulation methods. The review can provide basic theoretical knowledge and practical methods for 3D GPR road structural defections detection and provide guidance for digital transport infrastructure construction and road maintenance based on 3D GPR data.

Highway construction in China has witnessed remarkable achievements, with rapid growth in the national road network and continuous breakthroughs in the key technologies. This review aims to further enhance the influential level of pavement engineering in China, as well to promote its sustainable and high-quality development. The review systematically summarizes the current status, cutting-edge issues, and future development in pavement engineering. Specifically, it covers seven research topics:highway resilience evaluation and recovery, long-life pavement structures and materials, highway energy self-sufficiency, low environmental impact technologies, the genome of pavement materials and high-throughput computations, highway digitalization and intelligence, and highway intelligent inspection and high-performance maintenance. Focusing on the fields of green, resilience, intelligence, longevity, and traffic-energy interaction, the review identifies 20 critical research topics, including factors leading to highway disasters and their mechanisms evaluation and recovery of highway resilience, key technologies for enhancing highway resilience, full-scale tests for long-life pavement structures, technologies for extending the longevity of highway structures and functions, energy harvesting technologies, energy self-sufficient highways designs, environmental impact testing methodologies and evaluations; innovative materials for low-impact pavements, warm mix asphalt recycling technology, genomic studies on pavement materials, multiscale computation for pavement materials, research on the genome of pavement materials and high-throughput computations, digital modeling technologies, digital twin simulation technologies, data-driven technologies for highway maintenance operations, ground-penetrating radar detection technologies; lightweight detection of pavement performance, strategies for detecting and recovering pavement skid resistance, and high-performance preventive maintenance technologies. The review can provide guidance for the pavement engineering development in China, offering valuable reference for the researchers and practitioners in this field.

Intersections are bottlenecks in urban traffic operations and can cause congestion. Traffic control is a crucial strategy in regulating traffic flow by preventing and alleviating traffic congestion; it is also a cost-effective approach. The development of communication and automation technologies has led to a new type of mixed traffic, comprised of regular vehicles (RVs), connected vehicles (CVs), and connected and automated vehicles (CAVs). This has influenced the revolution of control objects, data environments, and control methods in urban traffic control. Beyond the challenges it brings, intersection traffic control provides opportunities for innovative developments in traffic control theory and methods. Intersection control in mixed traffic situations with CAVs is still in its infancy, but it has already become a hot topic both at home and abroad. According to the right of way, research on traffic signal timing, vehicle trajectory/path planning, and cooperative trajectory-signal control for mixed traffic control at an isolated intersection, in a corridor, and in a traffic network with shared facilities was first summarized; research on intersection control with dedicated facilities for autonomous driving was subsequently introduced, including CAV-dedicated lanes, CAV-dedicated links, CAV-dedicated zones, and dedicated bus rapid transit (BRT) lanes, which provide a separate right-of-way for CAVs. Even though current literature on intersection control in mixed traffic is still scarce, a comprehensive review on the topic has revealed a number of research questions that are yet to be addressed, including:the uncertainty in the driving behaviors of RVs; compliance rates of CVs; coupling of CV/CAV longitudinal and lateral trajectories; and spatial-temporal sparsity of CVs/CAVs. Future research directions are also herein presented, including:trajectory planning of CAVs in mixed traffic; controlling mixed traffic by CAV trajectory planning; and cooperative control of CAVs and mixed traffic. It is expected that these topics will drive forward innovation of traffic control in mixed traffic with CAVs.

To further enhance the strength of the field of automotive engineering and promote the development of automotive technology in China, this study systematically analyzes the academic research status, cutting-edge hot issues, latest research results, and future development prospects in the field of automotive engineering at both domestic and international levels from six aspects:automotive electrification and energy saving, intelligent and connected vehicles, vehicle dynamics and control, automotive NVH (noise, vibration, harshness) control and lightweight control, automotive electronics and electrical (E&E) and software technology, and automotive testing and evaluation technology. Automotive electrification and energy saving constitute key aspects of pure and plug-in hybrid electric vehicles, hydrogen fuel cell vehicles, extended-range electric vehicles, and energy-saving vehicles. Intelligent and connected vehicles are objectives of the research on intelligent driving environment perception technology, autonomous driving positioning technology, intelligent vehicle decision-making and planning, motion control technology, vehicle-road coordination, intelligent safety technology, Internet-of-vehicles safety technology, and intelligent cockpit and human-computer interaction technology. Vehicle dynamics and control are addressed by the research on brake-by-wire, steer-by-wire, suspension-by-wire, and chassis-by-wire cooperative-control technologies. Automotive NVH control and lightweight control involves the prediction and optimization of automotive aerodynamic noise, NVH control of pure electric vehicle systems, acoustic metamaterials and automotive structural vibration control, automotive noise active control, and automotive lightweight and collision safety technologies. Automotive E&E and software technology is addressed by the research on automotive E&E architecture, automotive software technology and OTA (over the air) upgrade, chip and system function integration, etc. Automotive testing and evaluation technology is addressed by the research on testing and evaluation technology of fuel vehicles, new energy vehicles, and intelligent and connected vehicles. This review provides a reference for further development of automotive engineering research in China, and guidance for the innovation in key technologies of the automotive industry.

Road traffic safety analyses and evaluation involve the entire process of road design, operation, maintenance, and reconstruction. The objective and content of road safety analysis and evaluation vary significantly in different stages and situations. Therefore, appropriate traffic safety analysis and evaluation methods must be selected for specific problems. To identify the characteristics and application of different road safety analysis and evaluation methods, in this study, four typical road traffic safety evaluation methods, including traffic crash statistics, vehicle operating speed, traffic conflict, and driving behavior analyses, were investigated. The characteristics and applications of different evaluation methods for data collection, characteristic indexing, modeling, and practical application, along with development trends, are summarized. In the literature, certain road traffic safety analysis and evaluation methods have formed relatively mature theoretical models and have been successfully applied to road traffic safety evaluations, such as methods based on traffic accident statistical and vehicle operating speed analyses. However, safety evaluation methods based on traffic conflicts and driving behavior have problems such as data sources and indicator consistency. There are still issues to be studied in the construction of data-sharing platforms, data integrity, temporal and spatial correlation of data, and analysis of data characteristics. The operational speed analysis method must be further studied in terms of precise data acquisition, model development and calibration, and zero-collision segment reliability threshold selection. In addition, traffic conflict analysis methods have unresolved problems in estimating conflict sample size statistics, estimating the construction of dynamic traffic conflict models, screening traffic conflict discrimination indicators, and assessing the impact of driver factors on traffic conflict modeling. Driving behavior analysis requires in-depth research on data collection consistency, collection time, and index consistency. Simultaneously, with the rapid development of intelligent and connected vehicles, more data related to road traffic safety operations may appear in the future, which may change our existing consensus on road safety analysis and evaluation methods and bring opportunities and challenges to the development of new road traffic safety analysis and evaluation methods. This review improves the understanding of theoretical systems of road traffic safety analysis and evaluation methods and enables the accurate evaluation of road traffic safety problems.

Distributed drive electric vehicles have a high degree of controllable freedom, fast response time, high chassis drive-by-wire integration, and compact vehicle structure, making them the best platform for realizing advanced vehicle dynamics control technology. The widespread use of wired control systems, such as steer-by-wire, drive-by-wire, brake-by-wire, suspension-by-wire, and different levels of driver assistance systems, such as anti-lock brake systems, lane keeping systems, adaptive cruise control, and lane change assist systems, result in redundancies and conflicts in the control of the vehicle chassis. The distributed driving format provides more possibilities for efficient and cooperative control between the multiwire control and auxiliary driving systems. This paper highlights the latest progress in distributed drive electric vehicle integrated control technology in terms of integrated control strategy architecture, longitudinal-lateral integration control, lateral-vertical integration control, longitudinal-vertical dynamic integrated control, longitudinal-lateral-vertical dynamic integrated control, fault-tolerant control, and chassis dynamics integrated control for distributed drive intelligent electric vehicles. It also provides an outlook on the development direction from multiple perspectives, aiming to provide a reference for developing high-performance integrated chassis control technology for distributed electric vehicles.

The existing macroscopic homogeneous model cannot reflect the impact of multiple scales, phases, and components on the mesoscopic heterogeneity of asphalt mixtures. Therefore, for the digital transformation of traffic infrastructure, it is important to construct a digital twin model of asphalt mixture performance from the mesostructure characteristics of asphalt mixtures to realize the accurate analysis, control, and optimization of pavement performance. Based on the digital twinning technology framework of asphalt mixture performance, in this article, the digital twin model of asphalt mixture performance is reviewed based on mesostructure characteristics, the digital extraction of the asphalt mixture mesostructure and the construction of the digital twin model are introduced, and digital characterization methods of image-based technology and artificial random generation are summarized. The evaluation methods for the asphalt mixture mesostructure composition are discussed, and the characteristic parameters of the mesostructure in the asphalt mixture digital twin model are summarized. The application of the digital twin model of asphalt mixture performance based on mesostructure characteristics in macro-meso performance correlation is described in detail. Finally, the development trends, model construction, parameter characterization, and performance evaluation of the asphalt mixture performance digital twin technology are discussed. The related work can provide valuable references and insights for the construction and application of digital twinning technology systems and technology platforms for asphalt mixture performance based on mesostructure characteristics.

The wide promotion and application of new-energy vehicles, especially electric ones, has resulted in huge charging demands. However, owing to the unreasonable planning of some charging stations, the spatial layout and service capacity of charging stations do not match the actual charging demands, and the coexistence of a low utilization rate of charging stations and inconvenient charging for users has become a significant problem in the electric vehicle industry. Reasonable location and deployment planning of charging stations is essential to optimize the usage and charging experience of electric vehicles, alleviate the “mileage anxiety” of users, improve the utilization rate of charging piles, and optimize the urban transportation and power grid network. This problem must also be solved in the field of intelligent transportation. To provide a more comprehensive overview of the problem, the development, types, and standards of charging stations are reviewed, and the charging demand estimation and prediction methods used on the “vehicle side” “ and “station side” are described. Then, charging station planning methods are analyzed from the perspectives of siting, sizing, and solutions. Research results on multiobjective, multiconstraint, and multiperiod covering models and flow-based models, as well as their application to simulation networks and actual city environments, are listed. Finally, existing problems in current research, including the accuracy of charging demand estimation and the match of charging stations to demands, are summarized. The coordinated planning for orderly charging guidance, new energy consumption, and “cut peak and fill valley” of the grid to make full use of the characteristics of distributed energy storage and flexible charging and discharging of electric vehicles in the era of vehicle-pile-road-grid interconnection is also discussed.

Path planning is an important part of automatic parking systems. It is the key to ensuring parking safety, shortening driving distance, and improving ride comfort. However, current automatic parking planning systems face certain technical challenges, such as narrow driving space, the presence of several obstacles, and difficulty in path search. Furthermore, the fixed radius of the search curve easily leads to discontinuous curvature at the path joints, increasing the difficulty of path-following control and the degree of tire wear. Accordingly, these factors complicate parking-path planning. Thus, in this study, a path optimization algorithm for automatic parking was developed based on hybrid A^* and the Reeds-Shepp curve with variable radius. Adjusting the curve radius can improve path-search ability and flexibility in complex scenarios. Moreover, a path optimization method based on the segmented Bezier curve and gradient descent was developed. Continuous multi-order derivatives were used to optimize the curvature of the searched path, and gradient descent was used to ensure path safety and avoid obstacles. Thus, discontinuous curvature changes at the point where a straight line meets an arc can be handled. The proposed parking planning method, which combines path search and path optimization, can effectively meet parking needs in complex scenarios. Finally, based on MATLAB and PanoSim virtual system, which was independently developed by Vehicle Controls and Intelligence Lab (VCI Lab), a joint simulation environment was developed to test and verify the proposed method under a variety of automatic parking conditions. The results demonstrate that the variable curve radius for global path search yields a shorter path that is easier to follow, and achieves considerable flexibility. The path optimization method based on the segmented Bezier curve and gradient descent can effectively eliminate sudden curvature changes, constrain path curvature, and ensure safe driving.

Auto-extraction of traffic safety information from traffic alarm reception data is of great significance for handling traffic crashes and improving traffic management. This paper established an auto-processing method based on a multi-task transfer learning algorithm that includes a text pre-training model as a shared parameter layer upstream, and a multi-task parallel processing method to automatically extract traffic safety information, types, and semantics. A total of 120 191 traffic alarm reception data items were collected over two years from a traffic control center in Suzhou Jiangsu, and a standard traffic safety information database was constructed with them. The experimental results show that the key information extraction method developed in this study can better extract the time, address, and license plate information from traffic alarm reception data. The performance of traffic crash classification achieved with the method proposed in this study is better than that of the conventional deep learning model: the classification accuracy is 93%. The traffic-safety semantic analysis method is based on local feature enhancement and focus on identifying the severity of crashes and rescue demands; its recognition accuracy is 87%. The results also demonstrate that the auto-processing method of traffic safety information proposed in this study has great portability and practicality.

To integrate and analyze pavement distress information inspected by non-destructive testing (NDT) and build a distress digital twin model, this study proposed a new building information modeling and geographic information system (BIM + GIS) framework for pavement distress integration and modeling. In this framework, surface and contour fitting methods were developed for modeling different types of surface diseases. Compared with traditional 2D picture patching or 3D fixed parameter modeling methods, the proposed method achieves the 3D modeling of distress. Another part of the framework proposes a 3D modeling method for concealed diseases in the pavement structure. Compared with traditional analysis and modeling methods, the proposed method transforms ground penetrating radar (GPR) data into distress models, efficiently integrates the model with the BIM platform, and lowers the threshold for the practical application of GPR data for road pavements. Based on the disease characteristics of two GPR images, the 3D reconstruction of suspected and local key areas of concealed diseases were identified and extracted. Finally, a digital twin model including global diseases was automatically constructed, which integrated the NDT data efficiently and transformed the disease in the virtual space. A heat map was drawn by GIS to feedback disease distribution and development and guide maintenance management. The experimental results show that the framework can complete the 3D digital transformation of full-field diseases. The modeling accuracies of the crack and pothole models were 80.13% and 98.17%, respectively. The model of road internal distress was compared with on-site coring to determine the position and type of disease accurately. For road surface diseases, the surface fitting model is suitable for scenes that require spatially distributed diseases. The model has network cracks and is loose, which makes it suitable for modelling diseases. The contour fitting model is suitable for scenarios that require simplicity and high efficiency. Its modeling effect is better for distresses that can expand, such as transverse cracks, longitudinal cracks, and potholes, owing to the simple physical information of the distress contour. The distress modeling based on the C-scan image is efficient for internal distress identification; however, it cannot distinguish distress types, which is suitable for the overall structural quality analysis of roads. Distress modeling based on B-scan images can subdivide distress categories; however, because it is limited by modeling efficiency, it is more suitable for the quality analysis of local key roads.

Cross-sea transportation infrastructure is characterized by large engineering scale, and high construction complexity. There are still some key problems in traditional operations and maintenance work, such as high comprehensive cost, low efficiency, poor accessibility, and large offshore safety risks. In order to realize the safe, reliable and efficient operation and maintenance of the Hong Kong-Zhuhai-Macao Bridge (HZMB), four common problems of the sea-crossing transportation infrastructure are summarized: low perception ability of service status, low utilization rate of monitoring information value, low efficiency of traffic risk active management and control, and low level of informationization and intelligent decision-making in operation and maintenance management. This paper analyzes the key technologies of intelligent operation and maintenance of the HZMB, and refines nine construction contents of intelligent operation and maintenance of the HZMB: the cross-sea traffic infrastructure operations based on 5th Generation Mobile Communication Technology (5G) and Internet of Things (IoT), the millimeter deformation monitoring and closed space positioning based on Beidou, the intelligent monitoring platform and big data fusion processing system for underwater structure, the inspection, testing, integration of emergency system based on unmanned aerial vehicle, the proximity detection and maintenance system for cross-sea bridges and tunnels based on the inspection robot, the service environment digitization and operational status monitoring and evaluation system for cross-ocean cluster facilities, the digital maintenance and management system for cross-sea cluster facilities based on the life-cycle hypothesis, the intelligent system of full-time traffic safety operation and rapid emergency response, the integrated operation and are summarized management platform for cross-sea cluster facilities. In addition, the overall technical architecture including device awareness layer, communication layer, basic resource layer, data support layer, business support layer, intelligent application layer and user interaction layer are also proposed. Lastly, the main technical features of intelligent operation and maintenance of the HZMB. The intelligent operation and maintenance of the HZMB can provide important reference for the operation and maintenance of other cross-sea transportation infrastructure.

An accurate and reliable evaluation is required prior to large-scale marketing of automated vehicles (AVs). However, as the system complexity increases and the operational designed domain expands, the evaluation methods employed for traditional vehicles can no longer meet the requirements of AVs. To solve the logical scenario phase safety evaluation problems, a method based on naturalistic driving data (NDD) is proposed in this paper. First, the construction process of logical scenarios based on NDD is developed, and the description parameters are analyzed. An NDD collection platform is used to gather related data, and the distribution of the different parameters is described by a Gaussian distribution. Next, the logical scenarios are discrete to obtain the concrete scenarios, and the tested algorithm is traversed in the obtained concrete scenarios using a joint simulation platform, which is a combination of PreScan, CarSim, and MATLAB. Hazardous concrete scenarios are clustered using Gaussian to build a dangerous domain. Finally, the scenario risk index, which considers both the parameter space of logical scenarios and the hazardous domain of the tested algorithm, is proposed, and evaluation examples are provided in leading braking and cut-in scenarios. The scenario risk index of the tested algorithm in the leading braking scenario is 0.409 8 and 1.08×10^-5 in the leading cut-in scenario. This means that the algorithm is more dangerous in the leading braking scenario, which is consistent with the visual results of the virtual test. By making a comparison of the scenario risk index and simulation test results, it can be found that the proposed evaluation method can be used to ensure that the safety of AV algorithms in the logical scenario phase is quantifiable and easy-to-operate, and verify that it is close to actual driving conditions.

Currently, China uses Vehicle Infrastructure Cooperation (VIC) as a main development path for autonomous driving. Relevant achievements are constantly updated and iterated, and the VIC industry is developing rapidly. Smart cars and intelligent roads are the core elements of the VIC industry, and an efficient synergy between the two is needed for a good foundation. VIC has the characteristics of wide coverage, a long industrial chain, and outstanding cross-border integration. Firstly, an overview of the VIC is described, and the key technologies and overall architecture are introduced. Secondly, China's basic advantages in the VIC industry are systematically presented. In contrast to America and developed countries in Europe, the advantages of state-owned land and road facilities, institutional advantages under the new national system, basic advantages of the manufacturing industry with completed categories, and potential market advantages of the world's largest autonomous driving market make a coordinated development of the VIC, the preferred way to develop intelligent transportation in China. Then, the problems in VIC development were analyzed in depth from several perspectives, such as planning, law, technology, and industry standards. Finally, countermeasures and suggestions for the development of China's VIC industry are proposed to accelerate development and promote the construction of intelligent transportation. The aim of this study is to achieve the development goal of a country with a strong transportation network, implement a dual-carbon development strategy, ensure national transportation information security, and promote a leapfrog development in transportation.

Vehicle movements can be highly random, and the driving style used becomes complex in urban traffic environments. To overcome the difficulties in accurately predicting vehicle trajectory in complex traffic environments, the Social Generation Adversarial Network (Social GAN) machine-learning model was used to develop a vehicle trajectory prediction algorithm named SIA-GAN. This developed algorithm was based on a spatial-temporal attention mechanism by considering a vehicle's speed, acceleration, course angle driving state, and shape size, and an interaction influence force field between the different vehicles was derived. Based on the magnitude of the interaction influence force that characterized each vehicle at the scene, different spatial attention weighing factors were assigned to the vehicles, along with a component of stressed "attention" that incorporated the information of vehicles having a greater impact on each other's driving pattern. The time attention mechanism was then combined to mine the time dependence of the vehicle under consideration in terms of the trajectory's feature vector during the observation period. To verify its effectiveness, the proposed algorithm was iteratively trained on an open-source dataset and compared with three trajectory prediction algorithms (long short-term memory (LSTM), Social LSTM, and Social GAN). The results show that SIA-GAN not only improves the convergence speed during training but also significantly reduces the average displacement error (ADE), final displacement error (FDE), average velocity error (AVE), and average course angle error (ACAE) when compared with other existing algorithms for trajectory prediction. At predicts 3.2 s, each of the aforementioned indexes decreases by 51.25%, 60.1%, 37.84%, and 13.75%, respectively, on average. The average reduction at predicts 4.8 s is 52.78%, 61.47%, 35.92%, and 9.57%, respectively. Thus, the proposed SIA-GAN trajectory prediction algorithm can accurately and effectively reflect complex spatial interaction characteristics between vehicles, enhancing the accuracy, rationality, and interpretability of trajectory predictions.

To widen our understanding of concrete-filled steel tubular (CFST) bridge towers and promote the application of CFST structures in cable-supported bridge towers, the engineering applications and general structure details of CFST bridge tower were first sorted out. The main problems existing in the design of CFST bridge towers were discussed, and the existing CFST bridge tower structure was optimized from the perspective of structure simplification and construction efficiency. Subsequently, the local buckling behavior of steel panels and the mechanical properties of CFST towers was reviewed, and the recommended design methods were provided. Finally, the technological characteristics and economy of the CFST bridge tower were compared with those of the traditional reinforced concrete bridge and steel bridge towers. The results showed that owing to the lack of an in-depth understanding of the common bearing mechanisms of steel and concrete, the relatively weak research on the local buckling theory of steel plates is subject to unilateral concrete restraints, unclear force transfer performance for the steel-concrete interface, complexity of the current structure of the CFST bridge tower, and poor construction efficiency. The optimized PBL-stiffened CFST bridge tower has a simpler stiffening structure, a steel-concrete connection structure, and steel panels that have a stronger restraint effect on the concrete without the need for reinforcement design; thereby reducing steel consumption, simplifying the steel structure manufacturing process and on-site installation process, and improving the level of industrial manufacturing and assembly construction of bridge towers. A structural design method for stiffened steel plates that considers the influence of local buckling and the calculation method of the bearing capacity of the CFST bridge towers is safer and more reasonable. CFST bridge towers have design flexibility, construction efficiency, and high disaster-bearing toughness. The cost of construction and maintenance is much lower than that of steel bridge towers, and it can economically compete with traditional reinforced concrete bridge towers, which has broad application prospects.

A general framework from semantic segmentation to geometric information extraction and integrated modeling was proposed based on LiDAR data to rapidly and automatically extract road geometric information and complete digital modeling. The local maximum and neighboring point mean features were concatenated as local features based on a fundamental foundation of spatial contextual features. Three-dimensional coordinates and radial distribution were combined to describe the global contextual features, and a semantic segmentation network was established. Additionally, the voxel grid filter and radius outlier removal methods were used to minimize the amount of point cloud data and remove outliers. The adaptive radius variable alpha-shapes method (VA-Shapes) was then employed to extract the road boundary based on semantic segmentation results. Furthermore, the geometric data of the road, including the road width, longitudinal gradient, and cross-slope, were obtained from the horizontal and vertical coordinates of the boundary. The in shape function and interpolation method were then applied to establish a digital elevation model. Subsequently, road routes were generated from the extracted road geometric information using Dynamo for Revit, and adaptive road components and various infrastructure components were constructed using Revit, developing a detailed digital road model. The Semantic3D dataset was utilized for training and testing to analyze and evaluate the extracted road geometric information. The overall accuracy (OA) of the proposed net is 95%, whereas the intersection-over-union (IOU) of segmented pavement is 97.9%, indicating that the proposed net could accomplish superior performance on semantic segmentation of point clouds. Compared with the traditional fixed radius A-Shapes method, the temporal complexity of the VA-Shapes method is low. In addition, the VA-Shapes method can efficiently extract the road boundary. The mean absolute errors between the extracted and manually measured geometric information are slight, demonstrating the effectiveness and accuracy of the proposed methods. The proposed process from semantic segmentation of the point cloud for geometric information extraction and building information modeling for digital modeling has the potential to build a digital model of a road in reverse, which is critical for the intelligent management of existing road infrastructures.

Electrified Road (e-Road), which is mainly characterized by Inductive Power Transfer (IPT), has been developing rapidly in recent years. It can provide dynamic wireless charging for moving electric vehicles, effectively solving the problems of long charging time and insufficient range of electric vehicles, and is an important reserve technology to support the development of electrification of road transportation.This review details the operating principles and performance characteristics of the IPT system and summarizes the charging performance parameters and TRL of existing e-Road test sections.On this basis, the main engineering problems and related research progress of e-Road are further analyzed from the perspective of infrastructure, including: ① an in-depth analysis of the impact of electromagnetic loss caused by high-frequency magnetic field through dielectric pavement materials on the charging efficiency of the IPT system and possible solutions; ② in view of the mechanical incompatibility between charging unit and asphalt pavement, the causes of the mechanical damage to the e-Road composite structure are inquired and potential optimization measures are proposed; ③ the life-cycle environmental performance of e-Road and conventional roads were assessed and compared, from which the environmental performance of e-Road is important and should be included in the full life-cycle sustainability framework for assessing the benefit of using electric vehicles.In addition, a comprehensive feasibility analysis of e-Road is conducted,in terms of policy support, safety, and price factors. An outlook is made on the future development of charging road infrastructure, including discussions over possible means to integrate e-Road with other new smart road technologies.

Based on the development of intelligent highways, typical logical and physical models of cooperative vehicle-infrastructure systems are summarized in this paper. Through the comparative analysis of demand between cooperative vehicle infrastructure systems on highways and other advanced transportation systems, intelligent highway systems can be better studied. After summarizing the overall architecture of domestic and foreign cooperative vehicle infrastructure systems on highways, this paper proposes a novel generation method for intelligent highway system architectures, including an overall architecture, intelligent classification, and data distribution mechanism for intelligent highways based on modularization. Additionally, according to the current mainstream application technology of cooperative vehicle infrastructure systems on highways, the status of research on emerging technologies driving the rapid development of intelligent highway systems is summarized. Research topics include high-precision vehicle navigation technology, advanced driver assistance system and vehicle bus technology, roadside unit optimization technology, heterogeneous network integration technology, network load balancing technology, network information security technology, multi-sensor fusion and cooperative awareness technology, user-centered scene adaptive information publishing technology, vehicle group cooperative autonomous driving technology, forecasting of transportation technology based on big data and artificial intelligence, lane-level active traffic management technology, and component-based application service development technology. Based on the characteristics of these 12 key technologies, recommendations regarding the application and promotion of intelligent highway system technology in the future are presented. In this study, typical scenarios of cooperative vehicle infrastructure systems on highways were analyzed, including traffic information services in broadcasting, active traffic management, syndrome traffic information services, automated vehicle dedicated lanes, and cooperative autonomous driving in vehicle platoons. Additionally, evaluation methods and relevant case studies on intelligent highway systems are summarized. Finally, the challenges and future development trends of intelligent highways are systematically analyzed and predicted. Highly trusted information interactions and intelligent collaborative control between people, goods, vehicles, roads, and clouds will provide a real-time and highly reliable traffic environment for autonomous driving on highways. The results of this study are significant as a reference for current and future technology research and the development of intelligent highways and engineering applications of intelligent highway systems.

To study the influence of dynamics and moisture content on the mechanical properties of embankment filling with coarse-grained soil, samples of coarse-grained soil under different moisture contents w were prepared. Dynamic stress with different frequencies f was applied to the samples, and then static triaxial compression tests were performed. The σ₀-ε₁ curves of the different samples were analyzed. According to the Janbu formula, the correlation between the dynamic deviator stress σ_d, moisture content w, load frequency f, and parameters n and K were explored. The fitting formulas for the initial deformation modulus E_i, ultimate deviator stress (σ₀)_ult, and controlled variables were established. An improved Duncan-Chang model was proposed for coarse-grained soil considering the influence of the dynamics and moisture content. Finally, verification tests were conducted to analyze the effectiveness of the revised model. The results show that the deformation of the samples changes from elastic to plastic when ε₁>0.5%. Under different controlled factors, the σ₀-ε₁ curves of the samples change significantly in the range of 0.5%<ε₁<2.0%. The tangent deformation modulus E_t also changes considerably within this range; the reduction rate reaches 58%-76%. When ε₁>2%, the change in E_t gradually decelerates and stabilizes. The samples exhibit strain-hardening characteristics during loading. As the moisture content, dynamic deviator stress, and load frequency increase, the deviator stress decreases when it reaches the same axial strain in the static triaxial tests. The initial elastic modulus E_i of the coarse-grained soil was related to the third principal stress σ₃. A linear relationship exists between lg(E_i/P_a) and lg(σ₃/P_a). The parameter n in the Janbu formula is not sensitive to controlled factors. In these tests, the value of n is stable at 0.78-0.80. The relationship between σ_d, w, f, and parameter K is nonlinear. Parameter K decreases with an increase in σ_d and w, while f shows an opposite trend. The power function can be used to establish the relationship between E_t and controlled factors, the relationship between (σ₀)_ult and controlled factors is the same. Validation tests show that the improved constitutive model in this study can effectively predict the σ₀-ε₁ curve of the coarse-grained soil filling of embankments with different σ_d, w, and f.

It is difficult to estimate the available capacity of lithium-ion battery efficiently and accurately after its decline.To address this problem,an increment capacity analysis method that does not rely on filtering algorithm is proposed to obtain the capacity decline characteristics of different types of batteries,and the estimation model of available capacity is built based on a data-driven approach.First,the shortcomings of low-pass filtering and wavelet filtering in obtaining increment capacity curves are analyzed respectively.Moreover,the patterns of increment capacity curves are compared for differential voltage values at 1,10,20,and 50 mV,respectively.Second,the moving variance algorithm is leveraged to evaluate the volatility of the increment capacity curve at different voltage differential values,and the volatility of the voltage value corresponding to the peak position of the curve is evaluated to determine the increment capacity curve with obvious and smooth peak characteristics.The peak value of the curve is extracted as the aging characteristic of the lithium-ion battery and the correlation between the aging characteristic and the aging state of the battery is examined using the Spearman correlation coefficient.Finally,two types of battery aging datasets prepared under different aging test conditions are employed for model validation.The results show that the established estimation model can effectively estimate the available capacity value within the full life cycle.Except for a few values,the relative error of the test results in the two data sets is within 2% for most of the relative error values.In dataset 1,the first 50% of batteries 1 and 3 are selected as training data and the second 50% are selected as test data,respectively,and the absolute error of training results is stable at approximately 0.05 A·h,and the absolute error of test results is approximately 0.04 A·h.These predictions are made for the full-life cycle discharge capacity values of batteries 2 and 3.The results show that the absolute relative error is restricted within at 2%.In dataset 2,the relative absolute error of the estimation results of the full-life cycle available capacity of batteries 5,6,and 7 is also less than 0.1 A·h (2%).The proposed model can make accurate estimations with less than 4% error when measuring the cycle of effective tracking of the capacity recovery phenomenon that occurs during the lithium-ion battery cycle.This result indicates satisfactory robustness and generalization ability.

An asphalt mixture is a multi-level and multi-phase granular material, in which the complicated aggregate characteristics, interface effect, and migration behaviors determine the segregation properties, compaction quality, and mechanical responses. Elucidating the mesoscopic mechanism of asphalt mixtures and providing a theoretical basis for the design optimization, construction control, and performance analysis of asphalt mixtures helps improve the durability of asphalt pavement. This paper reviews the outcomes of existing research on aggregate-asphalt systems. Asphalt mixtures used in paving, compaction, and service processes are regarded as different states of aggregate-asphalt systems with different degrees of freedom of migration. First, focusing on the aggregate characteristics, the effects of geometric morphologies and size of the aggregates on the performance of the asphalt mixture are analyzed; based on this analysis, the method for the calculation of composite geometric characteristics is described. Subsequently, considering the loose aggregate-asphalt system, the mesoscopic properties are described, including the contact-friction properties of the particle system and the interface interaction of the aggregate-asphalt system. Subsequently, the evaluation methods used for the segregation of asphalt mixture are summarized; the effect of particle migration on the segregation tendency of asphalt mixture is analyzed, and the forming mechanism of mixture segregation is discussed. Considering the compaction process of asphalt mixtures, the dynamic compaction properties of asphalt mixtures and the migration behaviors of particles are discussed; the influence of the geometric characteristics and the interface effect on the particle migration are analyzed, and the compaction process of the asphalt mixture is discussed based on the migration behaviors. Considering the compacted asphalt mixture, the spatial migration behavior of the aggregate-asphalt system is summarized, and the influence of the micromigration behavior on the mechanical strength properties of the asphalt mixture at the mesoscopic level is analyzed. Finally, the asphalt mixtures used in the transportation and paving, compaction, and service processes are defined as aggregate-asphalt systems with large-degree-of-freedom, small-degree-of-freedom, and micro-degree-of-freedom migration properties, respectively. The development tendency of interface behaviors and the migration dynamic properties of aggregate-asphalt systems are summarized, and prospects for further research are discussed. This review provides a reference for the basic theoretical research on the microscopic and mesoscopic characterization and mechanical response of asphalt mixtures, which will help improve mixture design, construction techniques, and maintenance quality.

Bio-oil is a green, environmentally friendly, and renewable resource with the potential to restore the physical and rheological properties of aged asphalt and improve road performance of recycled asphalt mixtures. To promote the in-depth research of bio-oil in the field of regeneration of aged asphalt materials, the source, preparation, and physicochemical properties of bio-oil were summarized, the regeneration mechanism of bio-oil on aged asphalt was discussed, the properties of bio-oil regenerated asphalt and bio-oil regenerated asphalt mixture were reviewed, and subsequent research interests were elaborated with regard to the existing deficiencies of bio-oil regenerated aged asphalt materials. The current research shows that pressed oil bio-oil has been most systematically studied for application in the regeneration of aged asphalt materials, followed by rich wood fiber plant-based bio-oil and animal manure bio-oil. However, all three types of bio-oil have broad application prospects in the regeneration of aged asphalt materials. The rich wood fiber plant-based bio-oil and pressed oil bio-oil play the role of “diluting” and “dissolving” in the regeneration of aged asphalt, which can regenerate aged asphalt in terms of both chemical balance and molecular structure repair. The animal manure bio-oil plays the role of “ dissolving” in the regeneration of aged asphalt, which achieves the regeneration of aged asphalt in terms of molecular structure repair by promoting the disintegration of asphaltene aggregates through its rich polar amide-based compounds. In addition, the rich wood fiber plant-based bio-oil and pressed oil bio-oil can be directly used for the regeneration of aged asphalt, while the animal manure bio-oil needs to be applied in combination with oil-rich regenerators to give full play to its regenerative effect. Among these bio-oils, the pressed oil bio-oil demonstrates better efficiency in regenerating aged asphalt materials. Future research is expected to be conducted as follows: establishing the correlation between the source, preparation process, and physicochemical properties of bio-oils to better screen and evaluate the applicability of bio-oils from different sources and preparation processes as asphalt regenerators; further exploring the application potential of rich wood fiber plant-based bio-oil and animal manure bio-oil in the regeneration of aged asphalt materials; focusing on the secondary aging of bio-oil regenerated asphalt and regenerated asphalt mixtures; and exploring the composite application of bio-oil in the field of regeneration of aged asphalt materials.

To explore the influence characteristics of automated driving takeover behavior in different scenarios. For drivers, automated vehicle and traffic environment, this study proposes a research framework for the automated driving test. Based on the driving simulation technology, an automated driving test platform was developed, and cases verified that this test platform can provide effective support for automated driving related technology tests. This study designed 18 freeways takeover scenarios with design elements of takeover request time, non-driving-task, scenarios, traffic flow and carried out driving simulation experiments to explore the adaptability differences of drivers from the subjective aspects. And from the objective aspect, the generalized linear mixed model was constructed to explore the influence of driver attribute factors (gender, age, driving age) and takeover situation factors (takeover scenario, takeover request time, no-driving-related task) and their interaction on takeover behavior. Statistical analysis results show that: ① There are statistical differences between male and female in trust and state perception of automated driving technology. Males have higher adaptability to automated driving technology than females. ② The driver's age and driving age have significant influence on the technology acceptance before and after the experiment, and there are statistical differences in the technology trust and state perception. Middle-aged people and elderly people, as well as people of middle and high driving age, have relatively high adaptability. ③ Different levels of factors lead to different takeover success ratio, takeover correct ratio and first control behavior. The generalized linear mixed model results show that: ① Takeover situation factors and their interaction have significant influence on takeover behavior indicators. ② There is an interaction between the driver attribute factor and the takeover scenario factor in the model. The study is based on driving simulation technology to develop an automated driving test platform, which is worth promoting. Besides, the study results can lay a foundation for further exploring the influencing mechanism of automated driving takeover behavior.

The continuous increase in bridge spans, innovations in structural systems and the application of lightweight high-strength materials have led to a decline in damping for long-span bridge structures. Consequently, it has intensified vibration sensitivity of bridges to dynamic loads, such as wind, vehicles, and earthquakes. This article is intended to summarize recent research findings on damping characteristics and identification methods for long-span bridges, and to promote progress in excitation technology and damping identification techniques. The review is organized in four aspects. Firstly, a systematic examination of the structural damping theories and a comprehensive review of the state-of-the-art of damping characteristics for long-span bridges are presented. Secondly, an extensive summary of the working principles, excitation systems, and application status of long-span bridge excitation technology are provided. Thirdly, the research progress of time-domain, frequency-domain, and time-frequency domain methods for long-span bridge damping identification based on ambient excitation is deeply analyzed, and the latest achievements in the areas of intelligent, automated, and uncertainty quantification of damping identification are systematically elaborated; in addition, the damping identification methods based on forced vibration and their application status are summarized. Lastly, the paper identifies directions for further research in four areas:damping theory, excitation technology, sensing systems, and identification methods. The findings from this review offer a theoretical foundation for the development of damping theory and identification method for long-span bridge structures. In addition, it is helpful to determine the suitable value of the structural damping for dynamic analysis of long-span bridges, and it may serve as a valuable reference for design and intelligent operation and maintenance of long-span bridges.

The traffic safety of ice and snow pavement has always been the difficult problem to be solved in transportation industry. Intelligent management of operation risk is an effective way to ensure traffic safety. The research progress of ice and snow pavement intelligent management at home and abroad was overviewed to promote the development of ice and snow pavement intelligent management technology and clarify the key issues and future direction. First, the impact of ice and snow on traffic were expounded, which reveals the relationship between people, vehicles, pavement and traffic accidents. It is clear that the root cause of the accident is the decline of pavement friction caused by ice and snow. Subsequently, the mechanism, influencing factors, prediction model and evaluation method of tire-ice-pavement friction on ice and snow pavement are discussed. Moreover, the operating principle and pavement engineering applicability of pavement snow and ice state sensing technology were compared and analyzed to put forward the technical requirements.Finally, the static and dynamic risk assessment and management methods are summarized. The static methods include meteorological method, historical accident data method and mechanical method, and the dynamic methods include vehicle dynamic method, traffic flow method and comprehensive risk assessment method. The comprehensive risk assessment method can represent the multi-factor dynamic change of road operation risk, which will be the way for intelligent management of ice and snow pavement in the future.

One of the most important factors that determine the passengers' acceptance of intelligent vehicles is the ride comfort. In order to improve the comfort of intelligent vehicles and help the design and optimization of intelligent driving control algorithms, this study performed a real vehicle ride comfort test based on the subjective perception of its passengers. In the test, the driver drove a traditional vehicle performing multiple lane-changing maneuvers. Data related to the 60 passengers' assessment of comfort and vehicle motion parameters were collected. Five vehicle motion parameters were selected for this study:the maximum lateral acceleration when changing lanes, the maximum lateral acceleration when returning, the maximum lateral jerk, the conversion amplitude of lateral acceleration, and the conversion frequency of lateral acceleration, and a binary logistic regression single factor analysis was used to evaluate their influence on the ride comfort. The receiver operating characteristic (ROC) curve analysis method was used to determine the comfort thresholds of these five vehicle motion parameters for passengers with different susceptibilities to motion sickness. A ridge regression analysis was used to determine the weights of the parameters' influence on the ride comfort. The results showed that:these five vehicle motion parameters have a significant impact on the ride comfort; the comfort threshold of the passengers susceptible to motion sickness is lower than the comfort threshold of passengers not susceptible to motion sickness; the maximum lateral acceleration when changing lanes and when returning, and the conversion amplitude of lateral acceleration are the main factors affecting the ride comfort when changing lanes. Finally, a comfort prediction model based on dynamic time warping(DTW)+ K nearest neighbor (KNN) algorithm was established according to the vehicle motion parameters and passenger physiological characteristics. The prediction accuracy of the ride comfort model is 84%, which can be used for the assessment of the comfort of intelligent vehicles control algorithms.

To promote the intelligentization of road infrastructure and realize the intelligent perception function of road paving structure, the research progress and development trend of road paving structure design system integrating perception characteristics were reviewed. First, the connotation of road paving perception information was clarified, including road environment perception, vehicle load perception, road surface function perception, and structure response perception. On this basis, the design idea of hierarchical perception of road paving structure was proposed. Aiming at the current imperfect design of road paving perception, a preliminary construction of a road paving structure design system that integrates intelligent perception characteristics was proposed. Finally, the problems existing in the design and application of road pavement structure integration with intelligent perception characteristics were identified, and its development and application were prospected. It is found that the perception design of road paving structure should comprehensively consider factors such as road grade, regional environment, service demand, project cost and development trend, and carry out hierarchical design based on the idea of “safety first, hierarchical progression”, and integrate the characteristics of intelligent perception. The road paving structure design system should cover the perception system, material design, layout combination, perception performance retention, calculation theory and design specifications and requirements. In the road paving structure design and application process that integrates the characteristics of intelligent perception, there were problems of the design connotation unclear, the implementation standards of perception technology not uniform, the performance maintenance technology of the perception pavement structure imperfect, and the perception data processing and application not standard. In the future, it has broad application prospects in the construction of the long-term performance observation scientific network of highways, the formulation of scientific and reasonable road maintenance decisions, and the research and development of long-life pavement design methods suitable for our country.

Bridges are crucial infrastructure for traffic and transportation. The inspection of bridge apparent defects is important for ensuring public safety, extending the lifespan of bridges, and identifying risks in a timely manner. They also contribute to improving the reliability and durability of bridges during their operational phases. In recent years, with the rapid development of technologies such as computer vision and artificial intelligence, machine vision has gradually emerged as a new approach for bridge apparent defect inspection. This study conducted a detailed analysis of relevant studies in recent years to review the key techniques for bridge apparent defect inspection based on machine vision, including inspection platform development, data acquisition, image processing, 3D reconstruction, defect localization, and defect parameter quantification techniques. By analyzing the inspection process of existing research, a technical framework for bridge apparent defect inspection based on machine vision was summarized, and the functions and connections between each process were analyzed. The above-mentioned review of key techniques and summary of technical frameworks provide a reference for researchers conducting inspection work on bridge structures. Finally, based on the different levels of automation in data acquisition and defect detection observed in existing studies, this study proposes a hierarchical classification for intelligent bridge apparent defect inspection based on machine vision. This classification includes six levels: manual inspection assistance, defect inspection and localization, partially automated inspection, globally automated inspection, high-degree automated inspection, and fully automated inspection. A comparison of existing literature reveals that although research has moved beyond the traditional stage of manual inspection, it still falls short of achieving fully automated inspection. Therefore, this field has strong research value and broad application prospects.

Accurate real-time perception of three-dimensional (3D) vehicle form is very important for many applications such as vehicle behavior analysis and traffic flow parameter estimation in intelligent transportation system (ITS) and autonomous driving. Among them, how to overcome the limitation of perspective projection and perceive 3D vehicle form by roadside monocular cameras is becoming one of the challenges in ITS. In order to solve this problem, we adopted deep convolution neural network (DCNN) to extract projection features, and combined geometric constraints in calibration space model to reconstruct 3D vehicle form from two-dimensional (2D) projection to 3D space. Firstly, based on our previous work, calibration space model was constructed for roadside camera to obtain the 2D-3D mapping matrix in perspective space. Then, based on the current popular deep network CenterNet, a simple and efficient DCNN, we designed the detection network of 3D vehicle form projection features with multi-scale feature fusion module integrated to optimize the detection of vehicles of different scales under perspective projection. At the same time, Gaussian convex hull heatmap was optimized to enhance vehicle feature detection. Prior geometric constraints in the enhanced loss function were also leveraged to accelerate the convergence of training. Finally, through the established geometric constraint model of 3D vehicle form, feature projection points were decoded from network outputs to construct complete 3D vehicle form information. The experiments were carried out on the public BrnoCompSpeed dataset and self-made dataset collected from roadside perspective. We manually labeled all the samples in the dataset which meet the requirements of the experiment and used data augmentation to simulate the variable camera perspective and environment. In the evaluation of experimental results, we evaluated the network detection results and the final constructed 3D vehicle form results, respectively. For the network detection results, the average precision (AP) of projection convex hull constructed by projection features was chosen as one of the evaluation metrics. When the 2D IoU threshold was set to 0.7, the AP obtained on the BrnoCompSpeed test dataset was 87.35% while the recall and precision were 87.39% and 90.78%, respectively. Besides, the ablation experiment was designed to prove the effectiveness of the network improved modules. For the 3D vehicle form results, we defined the metrics of 3D vehicle spatial localization, dimension, deflection angle and 3D IoU, and chose 3D IoU to verify the impact of multiple improved modules and different perspectives on the final AP. Finally, the average 3D IoU on BrnoCompSpeed test dataset reaches 0.738 and the frame per second (FPS) of the designed network is 27, which can achieve real-time requirements.

The asphalt binder is one of the most commonly used construction materials in pavement engineering, and is widely used in highways and urban roads in China. During pavement maintenance, asphalt materials exhibit aging behavior, such as hardening and brittleness, which significantly affects the operational quality and service life of the pavement. To promote the development of asphalt binder aging evaluation methods, this paper considers aging evaluation methods such as macroscopic performance testing, microscopic structure detection, and numerical simulation techniques. The macro- and micro-mechanisms that drive asphalt binder thermo-oxidative aging and ultraviolet aging were analyzed multidimensionally. The scales and application advantages of various evaluation methods were compared. The results show that current laboratory investigation methods for the thermo-oxidative aging of asphalt binders fail to accurately determine the long-term service durability. Thus, there is a great need for standardized laboratory research methods on ultraviolet aging. Thermal-oxidative aging is mainly due to high-temperature thermal decomposition that causes chemical bonds to break. Ultraviolet aging is triggered by the absorption of energetic UV light by groups in the molecule and their conversion from the ground state to the excited state, which leads to the breaking of chemical bonds. Macroscopic performance testing is a direct way to establish a link between asphalt binder aging and performance deterioration. The macroscopic performance testing methods and evaluation indicators for the two types of aging cannot be equated or replaced. Microscopic detection techniques, such as fluorescence microscopy, atomic force microscopy, scanning electron microscopy, Fourier-transform infrared spectroscopy, and transmission electron microscopy, can be used to qualitatively or quantitatively analyze the microscopic evolution process of asphalt binder aging and the aging response of modifiers. The results of molecular dynamics and micromechanical models correspond to the precise conclusions of microscopic investigations, and establish correlations with macroscopic mechanical tests to numerically characterize the aging process of asphalt binders. The results of this study can provide a reference for the microscopic evolution law and macroscopic performance properties of asphalt binders. In actual asphalt pavement maintenance, ultraviolet aging and thermal-oxidative aging occurs simultaneously. Multifactor composite aging simulations based on outdoor asphalt pavement aging monitoring should be the focus of laboratory asphalt binder aging simulation. The establishment of a multiscale asphalt aging dynamic simulation, which combines multilevel aging evaluation indicators and numerical simulation techniques, is the main research direction for the cross-applications of asphalt binder materials and numerical simulation.

To study the dynamic response mechanism of a vibratory roller during the process of subgrade compaction and provide theoretical support for the harmonic ratio index in continuous compaction control technology, based on the finite-element method, two elastic and elastic-plastic models were established, and the acceleration signal during compaction was simulated and analyzed. The effects of geometric nonlinearity, contact nonlinearity, and material nonlinearity on the distortion of the acceleration signal were studied. The generation mechanism of the harmonic and subharmonic components in the acceleration signal spectrum was illustrated, and their variations with soil parameters, such as modulus and cohesion, were analyzed simultaneously. During this process, the elastic model was used to determine the influence of geometric nonlinearity and contact nonlinearity caused by the change in contact area before jump vibration occurred. After the jump vibration, it was utilized to illustrate the impact of the contact nonlinearity caused by separation. Furthermore, an elastic-plastic model was established to reflect the effects of material nonlinearity. The results show that geometric nonlinearity has little effect on the distortion of the acceleration signal. Material nonlinearity and contact nonlinearity caused by contact area change lead to harmonic components in the acceleration signal, and contact nonlinearity caused by jump vibration induces subharmonic components. In addition, during compaction, the modulus and yield limit of soil increase at the same time, and the contact nonlinearity caused by the change in contact area is enhanced. However, the influence of material nonlinearity is weakened. Owing to the dominant elastic factor of soil in the middle and late stages of compaction, the plastic deformation is very small. Because of the influence of local disengagement, the second harmonic component increases during the compaction process. Finally, jump vibration not only causes subharmonic components, but also leads to a decrease in the harmonic components. Simultaneously, the acceleration waveform before and after the jump vibration does not change continuously but changes suddenly.

Effective measuring of full pavement performance is the key basis for comprehensive and systematic implementation of maintenance decisions. In this paper, we propose an automated method for pavement full-area performance measuring based on deep learning and virtual model. The method first generates pavement virtual entities based on UAV real measurement data modeling; then uses a deep semantic segmentation network to finely detect pavement performance from real measurement data; finally, the output performance feature detection results are matched and UV mapped with virtual entity data to obtain the positioning information of each performance in the virtual space and deploy them one by one to obtain damage state sensing model for the full area of the pavement. The results show that the average cross-merge ratio (MIoU) of the fully trained U-Net network reaches 0.86, showing an excellent segmentation accuracy of the pavement performance areas collected by the UAV. The pavement damage state sensing model established in this paper effectively senses 91 actual existing injuries in a pavement area of 236 m in length and 20 m in width, which enables a more systematic global characterization of pavement-wide injuries compared with the traditional two-dimensional inspection results, facilitating efficient inference of injury characteristics and location information, and realizing accurate and dynamic pavement service state assessment.

To verify the safety of the decision-making results of autonomous vehicles (AVs), a method for generating driving models with autonomous decision-making and interaction capabilities was proposed, and the driving models were as background vehicles (BVs) and used to build a self-evolution simulation scenario to test the continuous decision-making capability of AVs. First, based on reinforcement learning and a combination of inheritance and evolution ideas, different driving styles with autonomous decision-making and interaction capabilities were designed in this study. Second, in the model-building stage, three styles of driving models, namely, conservative, general, and aggressive, were generated and trained. The simulation training parameters for the general-style driving model were derived from the parameter distribution of a naturalistic driving dataset named highD to ensure fidelity. Finally, based on this, an aggressive-style driving model with significant aggressive features was designed and trained to enhance the complexity and testing effect of the self-evolution scenario. The results show that the distributions of parameters such as the car-following speed, distance headway, and lane-change moment time-to-collision obtained by using the highD dataset are in agreement with real data. An average similarity of 88% is observed between the general-style driving model generated and the corresponding real data, which is an improvement of 20.3% on the results obtained from the rule-based intelligent driver model (IDM). The proposed self-evolution scenario is seven times more testable than the baseline scenario composed of IDMs for the different driving models generated, as confirmed by the number of collisions in which the system under test is primarily responsible. Thus, self-evolution scenarios composed of the driving models designed and generated in this study can effectively support simulation tests for aiding the decision-making system in AVs.

To improve the accuracy and efficiency of the three-dimensional (3D) reconstruction of pavement macro-texture and to achieve high accurate evaluation of pavement anti-skid and anti-abrasive performances, a multi-view deep neural network has been proposed to perform pavement macro-texture reconstruction. First, a multi-view camera was used to collect pavement images with different perspectives. Then, a deep convolutional neural network was used to exact high-dimension features from each image, and the features were then mapped into a 3D matrix by a feature mapping unit. The 3D matrix was converted into a 3D voxel model by several deconvolution layers. Finally, the 3D voxel models in different views were combined by Bayesian rule, which was the 3D reconstruction result of pavement macro-texture, which was used to evaluate the anti-skid and anti-abrasive performances. The multi-view images and the 3D point cloud data were collected from 16 asphalt pavements by a multi-view camera and a 3D scanner, respectively. The two types of pavement data were combined to build a dataset to demonstrate the accuracy and stability of the proposed method. The test results showed that the multi-view deep neural network reconstructed the pavement macro-texture with the values of intersection over union (IoU) of 0.858 and 0.769 under 50 DPI and 70 DPI resolutions. The IoU results were not affected by the road surface material and background noise. The accuracy and stability of the proposed method were better than that of MVF-CNN, 3D-FHNET and stereo-vision methods. The 3D reconstruction results were used to evaluate mean texture depth (MTD) and dynamic friction coefficient (DFC) with the measured error is 6.82% and 7.28%, respectively, which meet the requirements of pavement performance detection. In addition, the test speed of MTD and DFC was 60 km·h^-1, which had high efficiency. The 3D model of pavement macro-texture can be used to build the highway digital twin in the future.

Electromagnetic waves emitted by ground penetrating radar are easily attenuated by the external environment. Complex underground municipal facilities in urban areas further increase the difficulty of cavity detection. Currently, the amplitude feature in the time domain cannot fully reflect the structure and dielectric parameters of the cavity, leading to cavity omission and misjudgment during automatic detection. To fully utilize the multi-attribute information of the ground penetrating radar signal, improve the accuracy and efficiency of automatic detection of urban cavities, the amplitude, frequency, and phase features of the reflected signal at a specific time can be extracted, and the feature fusion can improve the accuracy of cavity detection. First, the signal is converted from the time domain into the time-frequency domain using the Hilbert transform, and the amplitude, frequency, and phase profiles at a specific time are obtained by time-frequency domain calculations. Four single-feature datasets containing the original profile are obtained. Second, IA+IF, IA+IP, IF+IP, and IA+IF+IP are fused using the 2D wavelet transform method. In addition, the maximum fusion rule is used in the high-frequency domain while the mean fusion rule is used in the low-frequency domain. Finally, the YOLOv7 algorithm is used to train on these 8 datasets, and the performances of the training models are compared. The results showed that the models trained on the IA+IP and IA+IF+IP datasets had better performance than the model trained on the OP datasets; the model trained on the IA+IP datasets showed the best performance; its precision, recall, F1_score, and AP_0.5 rates were 5.0%, 7.6%, 7.8%, and 5.9% better than those of the original profile, respectively. The proposed method can depict other detailed information beyond the amplitude of the cavity and strengthen the reflection characteristics of the signal at the location of a cavity, which can improve the performance of the automatic detection algorithm.

With the development of automated driving, road tests have gradually been conducted for high-level automated driving vehicles in limited areas. Assuring and improving the safe driving capability of self-driving systems is a popular topic in current research, testing, and development. To reduce the traffic safety risk under mixed traffic conditions, a method of verifying and testing the compliance of automated driving vehicles with the same traffic rules is proposed. Aiming at the technical bottleneck of automatic semantic analysis of various traffic laws and rules, this paper proposes a two-stage digital model of normalization-logic traffic rules based on improved predicate metric temporal logic (MTL). Natural language traffic rules were transformed into logical codes constituting propositions, logical connectives, and time-series operators. In addition, digital traffic rules that can be understood, executed, and verified in automated driving systems were generated. A classification and grading system for traffic rule propositions was constructed. Furthermore, a set of traffic rule compliance verification algorithms based on high-precision trajectories of automated driving vehicles were proposed, a simulation test platform was built, and verification was performed in a highway traffic scenario. Theoretical analysis and test results show that improvements such as simplifying the proposition space, adding time series operators, and predicate logic words effectively improve the time representation ability of the original MTL framework, solve the problem of time series logic, and greatly improve the efficiency of the digital transformation of traffic rules. Additionally, the method is compatible with local traffic laws and future traffic law revisions. The proposed traffic rule compliance verification method and test platform can effectively test the ability of an automated driving system to comply with the existing traffic rules. These results are significant for improving the safety performance of automated driving systems and the level of hybrid traffic safety control in the future.

To improve punctuality, reduce delay, and solve bunching of urban buses, an optimization method for bus scheduling was studied based on a group-gathered passenger flow. The travel willingness, ride attributes, and passenger arrival regularity were all identified to feed the model. The bunching scene was described based on the vehicle carrying restrictions, delay at stops, arrival rate, and passenger alighting rate. Constraints such as punctuality, passenger flow demand, and control strategy were all considered in the development of a real-time mixed control strategy, which was adopted for multi-objective optimization of the minimum headway deviation and total passenger travel time. To prevent potential bus bunching, a scheduling method of bus bunching that meets the travel demands of periodic group-gathered passenger flows at bus stops is herein described. The method considers the uncertainty of passenger arrival rate, holding time at bus stops, and average driving speed across different sections. To solve the model, the difference in the view of bi-objective optimization was considered when the overtaking rule was used to reorder outbound vehicles in the bunching scene. To design the NSGA-II algorithm, the order relation was calibrated by the crowding distance, and a new population was obtained by the elite strategy to improve the crossover operator. The resulting Pareto solution set was then optimized based on the TOPSIS method. Finally, an actual bus line was used as an example to experimentally verify the accuracy of the model and its algorithm. The results show that the optimization model of bus bunching based on group-gathered passenger flow at stops systematically predicts the passenger riding attributes and vehicle carrying restrictions. An optimal scheduling scheme for vehicle holding and speed adjustment is obtained by the model, which allows for calculating a series of operational indicators, such as vehicle departure time, headway deviation, punctuality rate, passenger waiting time, and passenger travel time. Comparing the system before and after the model optimization, it is found that total headway deviation is shortened by 56%, reducing the total travel time of passengers by 11.7%, passenger average waiting time by 12.5%, and increasing the punctuality rate of bus stops by 24.1%. A reduction of 36 in the number of bunches is also observed. An experiment to verify randomness found that the average declined ratio of the two objective functions is 50.4% and 13.7%, which is a large reduction range and a good optimization effect. The results show that implementation of this model in real-life can greatly improve bus operational efficiency and effectively solve the problem of bus bunching; the solution is robust and reliable, and this method is both practical and feasible to implement.

Compared to scenarios under conventional ground motions, bridge structures subjected to across-fault ground motions have more complicated dynamic responses, larger seismic demands, and more severe seismic damage. With the development of super-infrastructure, such as the Sichuan-Tibet railway, an increasing number of bridges are facing the hazard of crossing potential faults; therefore, relevant studies are urgently needed. Based on the seismic damage of fault-crossing bridges, this paper first demonstrates the basic characteristics of near- and across-fault ground motions and introduces the relevant simulation methods. Then, studies on the analysis theory, numerical simulation, and model tests of fault-crossing bridges are reviewed, and coping strategies against the fault-crossing effect are summarized. The results show that the current simulation methods of across-fault ground motions are generally reasonable and feasible, but certain application conditions exist and improvements are needed; the studies on fault-crossing bridges mainly focus on middle- and small-span girder bridges, and the seismic responses and damage modes of bridges are affected by multiple factors, including fault type, fault-crossing location, fault-crossing angle and permanent displacement. The seismic isolation technique and the anti-collapse displacement restriction measurements for fault-crossing bridges have achieved preliminary improvements. Finally, development orientations for future studies on fault-crossing bridges are proposed in this study. In particular, more accurate and efficient simulation techniques for across-fault ground motions need to be developed; shake table array tests and numerical studies on long-span fault-crossing bridges need to be conducted, the difficulties with the experimental technique of shake table array tests for fault-crossing bridges need to be solved, the influences of ground motion parameters on different types of fault-crossing bridges need to be explicitly pointed out, and seismic isolation, anti-collapse, and resilient techniques should be further developed and experimentally validated. This study can provide a reference for the simulation of across-fault ground motions and the analysis and design of fault-crossing bridges.