Current Issue

15 January 2026, Volume 45 Issue 1 Previous Issue   
---
Intelligent Traffic Infrastructure
Traffic & Transportation+Artificial Intelligence
Modern Traffic Equipment


For Selected:

Download Citations EndNote Ris BibTeX

Toggle Thumbnails

Intelligent Traffic Infrastructure

Select

Comparative Study on Calculation Models of the Transition Section Length of Parallel Acceleration Lane on Freeway SHAO Changqiao, ZHANG Dongyu 2026, 45(1): 1-6.  DOI: 10.3969/j.issn.1674-0696.2026.01.01 Abstract ( )   PDF (1411KB) ( )   The transition section of parallel acceleration lane plays a critical role in ensuring the operational efficiency and safety of freeway merging areas. Existing models for calculating the length of transition sections yield significant differences in results due to varying considerations. Based on this, a comparative study on multiple calculation models for transition section length of parallel acceleration lane was conducted. According to the measured data and the literature at home and abroad, theoretical calculation results of different models were given. Taking the average operating speed of merging vehicles and the number of traffic conflicts of mainline vehicles as indicators, VISSIM simulation was employed to analyze the rationality of transition section length corresponding to each model. The research results indicate that the vehicle average speed corresponding to the transition rate model is the highest, the traffic conflict number of the lateral lane-changing model is the least, and the lane-changing driving distance model performs optimally in terms of both average speed and traffic conflict number. Finally, it is recommended that when the design speed of the mainline is 120 km/h, 100 km/h and 80 km/h, the lengths of the transition sections of parallel acceleration lanes are respectively set at 150 m, 130 m and 115 m. References | Related Articles | Metrics

Select

Estimation Method of Mean Texture Depth for Cement Concrete Pavements YE Juntao, ZHANG Dawei 2026, 45(1): 7-14.  DOI: 10.3969/j.issn.1674-0696.2026.01.02 Abstract ( )   PDF (9773KB) ( )   To address the issues occurred in the current sand patch method for measuring mean texture depth (MTD), a novel MTD estimation approach based on 3D point cloud data was proposed. Initially, 300 sets of cement concrete road surface samples with varying degrees of wear were scanned to collect point cloud data, which were then processed through horizontal correction, outlier removal, and Gaussian filtering. Subsequently, the height of the sand patch top surface was determined by using the second derivative of the cumulative percentage of point cloud heights, and subarea division and polynomial surface fitting techniques were employed to generate the sand patch top surface. Finally, the MTD was estimated by calculating the envelope volume between the sand patch top surface and the road surface through the Monte Carlo method. The results indicate that the random sample consensus algorithm （RANSAC）can effectively perform horizontal correction on cement concrete pavement. The combination of outlier removal with a threshold of 3σ and Gaussian filtering with a kernel standard deviation σ=1 and a search radius r=2 mm can accurately remove outliers in the point cloud data while preserving crucial texture details of the road surface. The estimation results for cement concrete road surfaces with different levels of wear exhibit strong correlations with the measured values, with R2 values of 0.957 2, 0.907 8, and 0.919 1, respectively. The overall estimation accuracy is high, and the relative error (-8.06% to 4.75%) is significantly lower than that of existing methods. References | Related Articles | Metrics

Select

Modification of Prediction Model for Carbonization Depth of Bridge Concrete under Southeastern Coastal Atmospheric Environment ZHANG Jianzhong1, ZHOU Xingyu2, LIN Xuechun1, ZHUO Weidong2, GU Yin2 2026, 45(1): 15-22.  DOI: 10.3969/j.issn.1674-0696.2026.01.03 Abstract ( )   PDF (2407KB) ( )   To investigate the laws of natural carbonization depth and their influencing factors of existing concrete structures in the atmospheric environment of southeastern coast of China, field surveys on the ambient temperature, humidity, CO2 concentration, concrete strength, and carbonization depth of 142 highway concrete bridges located in the southeast coastal area of Fujian Province with different service ages were conducted. Based on 142 sets of data measured on-site for over six months, the relationship between the carbonization depth of concrete and various influencing factors in natural environment was explored. It is found that: the compressive strength of concrete has the greatest impact on the natural carbonization depth, followed by the age of the bridge and the water-cement ratio. The ambient relative humidity, temperature and CO2 concentration have a relatively minor impact on the natural carbonization depth due to their small variation amplitude. Four kinds of carbonization depth prediction models (including FIB model, Chinese specification model, Portuguese LNEC model, Japanese JSCE model) were tested by use of 142 sets of on-site measurement data. The findings demonstrate that the carbonization depth prediction model proposed by China’s current specification exhibits the smallest deviation from the on-site measured data. Utilizing multivariate nonlinear regression analysis method, the influence coefficients of the Chinese specification model were further modified, and a concrete carbonization depth prediction model suitable for the atmospheric environment of the southeast coast of China was proposed. The mean square error (MSE), root mean square error (RMSE) and mean absolute error (MAE) of the proposed model are 6.75, 2.60, and 1.85, respectively, which can well reflect the natural carbonization degree of concrete under the atmospheric environment of the southeast coastern of China. References | Related Articles | Metrics

Select

Axis Extraction and Inclination Inspection of Bridge Underwater Pile Foundation Based on Sonar Point Cloud LIU Ning1, ZHANG Jiewu2, YAN Jing3, ZHU Yanjie3 2026, 45(1): 23-29.  DOI: 10.3969/j.issn.1674-0696.2026.01.04 Abstract ( )   PDF (15579KB) ( )   Underwater structures located in deep and turbid waters, facing challenges such as poor accessibility and considerable detection difficulties. To evaluate the stability of bridge underwater foundation and promote the development of inspection technologies, a method for extracting the axis of underwater pile foundation and automatically detecting the inclination was proposed, which was based on underwater sonar point cloud data. The principal component analysis (PCA) method was employed to determine the normal vectors of the point cloud slices of the pile foundations. Subsequently, projection and geometric analysis techniques were combined to extract the spatial coordinates of the pile foundation axis feature points. Then, the RANSAC algorithm was used to fit the axis of the underwater pile foundation, thereby enabling the calculation of pile foundation inclination. The effectiveness of the proposed method was verified through the point cloud models of underwater pile foundation generated by numerial simulation. The research results show that the mean absolute error between the calculated inclination angles of pile foundations and the true values is 0.031°, which indicates that the proposed method can accurately extract inclination angles of pile foundations from point cloud data, possessing good engineering application value. Furthermore, 3D point cloud data of underwater pile groups of certain two Yangtze River bridges are obtained by utilizing the three-dimensional sonar measurement system, by which the deformation detection of actual bridge underwater pile foundations is realized. References | Related Articles | Metrics

Select

nvestigation on Non-uniform Temperature Field Considering Solar Radiation of Steel Truss Cable-Stayed Bridge with Two-Layer Decks YANG Yibo1, ZHANG Yuanhao2, ZHAO Shichao1, GUO Zhiyuan1, ZHU Jinsong2 2026, 45(1): 30-38.  DOI: 10.3969/j.issn.1674-0696.2026.01.05 Abstract ( )   PDF (12126KB) ( )   In order to investigate the non-uniform temperature field of the steel truss cable-stayed bridge with two-layer decks under solar radiation, a numerical simulation method of temperature field of the steel truss cable-stayed bridge with two-layer decks under solar radiation was developed, which considered geographical meteorological factors and the occlusion relationships between various components. The method was validated through experimental data. Taking Jinan Yellow River Highway Bridge expansion project, with a main span of 488 meters in Shandong Province, as a case study, a three-dimensional heat conduction model was established by ABAQUS software. The effects of occlusion on the temperature distribution of bridge and the temperature distribution patterns along the horizontal and vertical directions of the steel truss surface under typical seasonal conditions were analyzed. The results show that the temperature time-history simulated by finite element agrees well with the experimental values, validating the accuracy and reliability of the numerical simulation method. Under solar radiation, the surface temperature field of the steel truss cable-stayed bridge with two-layer decks exhibits significant time-varying characteristics and spatial non-uniformity. Due to occlusion from the upper deck, the vertical temperature difference of the steel truss structure can reach 23.0 ℃ in summer, and the occlusion effect on the temperature distribution of the bridge surface is non-negligible. References | Related Articles | Metrics

Select

Influence of Shoal Regulation in Mountainous Waterway on the Upstream Behavior of Grass Carp HU Jiang, LI Jianshu, LI Geng, LIN Siyun, ZHAO Tingjun, CEHN Shuang, CHEN Xin 2026, 45(1): 39-44.  DOI: 10.3969/j.issn.1674-0696.2026.01.06 Abstract ( )   PDF (7948KB) ( )   In the regulation of shoal waterways in mountainous rivers, spur dikes are usually built on the side beach, or the river channel during medium-low water periods is narrowed along the dike, increasing the scouring capacity of the water flow to maintain the stability of the waterway. However, the beach slow-flow area in the shallow reach is often an important habitat for migratory fish such as grass carp, and the waterway regulation buildings have changed the shape of the beach, forming complex flow patterns such as rapids and large-scale eddies, thus affecting the migration of fish. To address this issue, a method combining high-precision flume experiments and three-dimensional flow mathematical models was adopted to study the hydrodynamic conditions of beach and the variation characteristics of grass carp upstream migration behavior before and after the construction of spur dike in shallow river. The results show that after constructing a spur dike on the beach, the river channel cross-section contracts, and its impact on water flow and the upstream migration of grass carp decreases with the increase of the submergence depth of the spur dike. When the spur dike is not submerged, the head of the spur dike is in the high velocity area, so it is extremely difficult for grass carp to pass through, and the shelter area behind the spur dike becomes a gathering habitat. When the spur dike is submerged shallowly (H/D=1.125), the success rate of grass carp upstream migration decreases by about 20%. When the spur dike is submerged deeply (H/D=1.875), the success rate of grass carp upstream migration decreases by about 4%. Generally, spur dikes used in waterway regulation projects are all short spur dikes (the length is less than 1/3 of the river width), and the flooding degree of spur dikes during grass carp migration is generally 1.200~1.500, so the impact of spur dike on the success rate of grass carp upstream migration is below 20%. References | Related Articles | Metrics

Traffic & Transportation+Artificial Intelligence

Select

Hierarchical Variable Speed Limit Control Strategy Based on Safety Goal JIAO Pengeng1, ZHANG Yao2, BAI Ruyu1, ZHANG Yao1 2026, 45(1): 45-2.  DOI: 10.3969/j.issn.1674-0696.2026.01.07 Abstract ( )   PDF (4304KB) ( )   To enhance driving safety, traffic efficiency and environmental performance of highway merging areas, the variable speed limit control strategy was investigated. Based on real-time traffic flow status obtained by the detector, with the shortest travel time and collision time as safety objectives, the optimal speed limit value was determined with the consideration of the traffic demand in different areas, and a hierarchical variable speed limit control strategy was established. By integrating prioritized experience replay and the ε-Greedy strategy into the double deep Q network (DDQN) algorithm, the optimized double deep Q network (OPDDQN) algorithm was proposed to improve the utilization rate of critical samples and enhance the stability of exploration strategy by training , thereby obtaining the corresponding control strategy. Relying on the SUMO platform to build a simulation environment, the effectiveness of the proposed strategy was verified under low, medium and high traffic flow conditions. The research results indicate that compared to the DDQN algorithm, OPDDQN reduces training time by 33%. In contrast to the no-control scenario, the proposed strategy reduces collision risk, travel time, and fuel consumption by 39.18%, 48.30%, and 34.48%, respectively, while increasing average speed by 67.05%. References | Related Articles | Metrics

Select

Nonlinear Impact of Built Environment on the Combined Travel Mode of Ride-Hailing and Metro PAN Yiyong, WANG Congwei, SONG Wenchao, JIA Xilai 2026, 45(1): 53-62.  DOI: 10.3969/j.issn.1674-0696.2026.01.08 Abstract ( )   PDF (9005KB) ( )   To explore the nonlinear impact of the built environment on the demand for ride-hailing and metro combined travel, firstly, based on the trajectory data of ride-hailing orders, the data of ride-hailing orders that were picked up and dropped off within a 500-meter buffer zone of metro stations were identified and screened out. Secondly, a gathering buffer zone for the ride-hailing and metro combined travel was established. According to the “5D” principle, built environment variable indicators were constructed, and the combined travel data and variable data were mapped to grids. Finally, the XGBoost model was used to analyze the nonlinear impact of the built environment on ride-hailing and metro combined travel and identify the key factors. The research results show that the fitting effect of the XGBoost model is overall superior to that of the other selected models. The distance to the city center and population density are the key factors affecting ride-hailing and metro combined travel, whose relative importance exceeds 45% in the case of pick-up combined travel and reaches 40% in the case of drop-off combined travel. The impact of the built environment on different combined travel modes of ride-hailing and metro shows significant nonlinear characteristics and corresponding threshold effects. The distance to the city center, population density, and road network density have negative impacts on combined travel, while commercial facility has a positive impact. The number of financial facilities is at a critial threshold of 16. Values below this threshold indicate a positive impact, while values above this threshold indicate a negative impact. In addition, the nonlinear characteristics exhibited by bus stops during different time periods for combined travel indicate a complex dynamic relationship between ride-hailing services and urban buses. References | Related Articles | Metrics

Select

Traffic Flow of Motorized Vehicle and Non-motorized Vehicle Based on the Integration of Survival Analysis Method and Improved NaSch Model LI Bing1, 2, YANG Jingfeng1, 2, LU Dazhi2, 3, YIN Juyuan1, 2, BAI Wenqiang1, 2 2026, 45(1): 63-74.  DOI: 10.3969/j.issn.1674-0696.2026.01.09 Abstract ( )   PDF (15994KB) ( )   Constrained by road infrastructure and the growing number of non-motorized vehicles, the coexistence of motorized and non-motorized vehicles has become a typical feature of urban roads in China. The modeling of this mixed traffic flow is highly complex due to the shared use of lanes. To address the shortcomings in existing research regarding the analysis of dynamic changes in lateral interference, the influence of various traffic factors on lateral interference in mixed traffic was investigated, which was based on the accelerated failure time (AFT) model in survival analysis method. Secondly, the IKKW model was introduced to optimize the NaSch model. Subsequently, the survival analysis method was integrated with the improved NaSch model to construct a mixed traffic flow model for motorized vehicles and non-motorized vehicles that quantified lateral interference. Finally, the characteristics and spatial distribution of mixed traffic flow were analyzed through simulation and actual data analysis. The research results indicate that mean the critical lateral distance between motorized and non-motorized vehicles follows a Weibull distribution. Motor vehicle speed, non-motorized vehicle load status, and traffic volume significantly affect the lateral interference probability, which increases with the increase of motor vehicle speed, non-motorized vehicle speed and speed difference, but decreases with the increase of non-motor vehicle load status and traffic volume of motorized and non-motorized vehicles. When the density of non-motorized vehicles is low, motorized vehicles are mainly subject to lateral interference, leading to a decrease in the critical density of traffic flow. When the density of non-motorized vehicles is high, motorized vehicles are mainly subject to longitudinal interference, leading to an increase in the critical density of traffic flow. The free-flow speed of motorized vehicles is synchronized with that of non-motorized vehicles, and the number of lateral interference events is significantly higher under low-density conditions than under high-density conditions. The actual data shows that mean the absolute percentage errors (EMAP) of non-motorized vehicle flow and lateral interference frequency are 16.98% and 8.72% respectively, indicating that simulation can effectively simulate complex mixed traffic environments. References | Related Articles | Metrics

Select

Reversing Variable Lane Design and Signal Timing Optimization Considering Saturation Time Non-stationarity GAN Zuoxian, LIU Yaxin, ZHAO Ruijia 2026, 45(1): 75-85.  DOI: 10.3969/j.issn.1674-0696.2026.01.10 Abstract ( )   PDF (5060KB) ( )   Reversible variable lane is considered an effective traffic organization method to alleviate intersection congestion caused by the imbalance between the supply and demand of left-turning vehicle flow. Efficiency index calculation models applicable to different saturation levels were introduced to establish a dual-objective optimization model for reversible variable lane design maximizing traffic capacity and minimizing delay. A hybrid algorithm combining with multi-objective evolutionary algorithm (MOEA) and non-dominated sorting genetic algorithm II (NSGA-Ⅱ) was employed to solve for the timing and spatial geometric parameters required for calculating the setting of reversible variable lanes. The effectiveness of the proposed optimization model under three different traffic flow loads, namely low, medium and high, was validated by VISSIM simulation. Meanwhile, based on the time non-stationarity of traffic flow load during peak hours, the optimal recommended scheme for reversible variable lanes was determined under three traffic flow ratio scenarios: continuous, concentrated, and balanced. The results show that under the combined effect of the setting of reverse variable lanes and the signal timing parameters obtained from the optimization model, the average vehicle delay at intersections under three different traffic flow loads is respectively decreased by 13.14%, 16.38%, and 23.34%, the number of stops is respectively decreased by 16.86%, 12.24% and 33.19%, and the queue length is respectively decreased by 18.62%, 24.43%, and 32.41%. Under fluctuating traffic flow conditions, the impact of time non-stationarity on the selection of the balanced flow scheme is relatively small. However, continuous and concentrated traffic flows characterized by prominent oversaturated traffic flow require variable lanes of sufficient length to assist vehicle traffic. In actual timing, it is necessary to comprehensively consider the proportion of arriving traffic volume. Simply considering the cycle and pre-signal design scheme under the maximum flow background will exacerbate the overall delay at intersections. References | Related Articles | Metrics

Select

Prediction of Driving Risk Level Based on Extreme Value Theory CUI Mengmeng, ZHU Yongming 2026, 45(1): 86-85.  DOI: 10.3969/j.issn.1674-0696.2026.01.11 Abstract ( )   PDF (2419KB) ( )   To address the problem of assessing driving risk levels for intra-city logistics distribution vehicles, extreme value theory and floating car data were employed to carry out the study. Firstly, six types of risky and safe driving behaviors were extracted from distribution data, and their velocity safety entropy was calculated. Secondly, a generalized Pareto model for velocity safety entropy was established based on extreme value theory. The threshold (0.189 8~0.199 8) was determined by combining the graphical method with sliding time window method, and the model parameters were solved via Bayesian estimation. The proposed model could be used for predicting and assessing driving risk levels. The reproducibility level theory was adopted to validate the effectiveness of the proposed model. The research results show that the approximate linear goodness-of-fit for the minimum threshold model and the maximum threshold model are 0.827 3 and 0.855 9, respectively, with the maximum threshold model demonstrating superior predictive performance for risky driving behaviors. The proposed model establishes a correlation between non-risky driving behaviors and risk levels, enabling the prediction and assessment of risk levels based on non-risky behavior data, thereby expanding assessment methods. It provides support for establishing an operational risk assessment and early warning system for intra-city distribution vehicles, discriminating risky driving behaviors and enhancing driving safety. References | Related Articles | Metrics

Select

Audio-Visual FusionDetection Method of Traffic Volume Based on Cross-Modal Multi-head Attention Mechanism MA Qinglu1, WU Feifei1, WU Yuechuan1, ZHANG Li1, ZHANG Geng2 2026, 45(1): 95-94.  DOI: 10.3969/j.issn.1674-0696.2026.01.12 Abstract ( )   PDF (20036KB) ( )   Aiming at the problem that traditional visual or audio signals cannot fully capture the detailed information in time domain and frequency domain in traffic volume detection, a traffic volume audio-visual fusion detection method based on cross-modal multi-head attention was proposed. In the proposed method, a cross-modal traffic volume detection model that spanned both audio and video modalities, was established to obtain high-quality traffic visual modal representation and sound modal representation and efficiently fuse them. Firstly, the Res2Net and DCNv3 networks were employed to extract features from audio and video data, while the bi-directional long short-term memory (BiLSTM) network was used to process time series features. The complex behavior sequences in audio and video were respectively analyzed to obtain rich and coherent traffic information description. Secondly, in the cross-modal fusion, cross-attention was integrated with multi-head attention, and multiple subspaces were used to combine the output to perform multi-head attention cross-modal fusion. Finally, the joint application of cross-entropy loss and consistency loss enhanced the coordinated analysis of information from different modalities, ensuring consistent performance of multi-modal data in classification and recognition tasks. Experimental results demonstrate that in the traffic volume detection scenario, the proposed method outperforms the average vehicle detection accuracy of single audio, video, and the fusion method in AVSS (audio-visual speech separation, AVSS) by 2.57%, 1.70%, and 0.95%, respectively; the average vehicle classification accuracy is improved by 4.72%, 1.78%, and 1.62%, respectively; the average detection accuracy of overall traffic volume is enhanced by 4.41%, 2.96%, and 1.46%, respectively. Furthermore, the performance remains stable in four distinct scenarios. References | Related Articles | Metrics

Modern Traffic Equipment

Select

Method for Extracting Motion Information from Vehicle-Mounted Visual Images LIU Ping1,2, WANG Shuohan1,2, ZHANG Yikang1,2, ZHOU Zilong1,2 2026, 45(1): 106-105.  DOI: 10.3969/j.issn.1674-0696.2026.01.13 Abstract ( )   PDF (11675KB) ( )   Motion object detection is an important research topic in the field of computer vision. Motion information is defined as the pixel position corresponding to the motion object in the image. However, in the context of autonomous driving, accurately extracting motion information is challenging due to changes in the image background caused by the motion of the onboard camera itself. Therefore, a motion information extraction model based on sparse optical flow estimation and deep learning was proposed to overcome the impact of background changes and detect motion information in the environment. The optical flow extraction module initially obtained global sparse optical flow by Shi-Tomasi corner detection and Lucas-Kanade (L-K) sparse optical flow estimation. The motion information discrimination module inferred suppression signals by inputting image depth information and sparse optical flow into a Transformer neural network, thereby suppressing the impact of background motion and extracting accurate motion information. The results show that the proposed method can extract motion information from images with an accuracy of 92%, which can be utilized for detecting moving targets for autonomous vehicles. References | Related Articles | Metrics

Select

Design of Marine Rim-Driven Thruster Based on Flux-Switching Motor MA Zhaosheng, YANG Rongfeng, CHEN Yao, LIAO Weiqiang, YU Wanneng 2026, 45(1): 113-119.  DOI: 10.3969/j.issn.1674-0696.2026.01.14 Abstract ( )   PDF (5777KB) ( )   Aiming at the issues of sand and gravel easily entering the air gap of the integrated rim thruster of ship, as well as the difficulties in disassembly and maintenance, an integrated propeller structure which was easily disassembled and assembled based on flux-switching motor was proposed. Firstly, the working principle of the flux-switching motor was analyzed, and the problems faced by the application of flux-switching motors in rim-driven thruster were discussed. The design idea of propulsion motor was put forward to reduce the influence of rotor tooth height by increasing the number of motor poles. Secondly, the structure design of the integrated thruster was discussed in detail. A kind of integrated propeller structure which was convenient to disassemble and assemble was put forward through the ring winding, concentric circle design and side slot design. Finally, the working performance of the propulsion motor was verified by Maxwell simulation. In particular, the difference of electromagnetic characteristics of three winding structures were compared, and the effectiveness of the propeller with this structure was demonstrated. References | Related Articles | Metrics

Select

Lithium Battery Remaining Life Prediction Based on Dual Strategy Optimization of VMD-HO-LSTM YANG Pengpeng1,2, ZENG Shenghao2, XUE Hai2, BAI Yongliang2 2026, 45(1): 120-128.  DOI: 10.3969/j.issn.1674-0696.2026.01.15 Abstract ( )   PDF (3543KB) ( )   In order to solve the problem of insufficient prediction accuracy of state of health (SOH) of lithium batteries, a model based on variational mode decomposition and the hippopotamus algorithm optimized long short-term memory (VMD-HO-LSTM) neural network was proposed to predict the remaining life of lithium batteries. Firstly, in order to eliminate the false capacity signal of lithium battery, the variational mode decomposition (VMD) method was used to decompose the capacity of lithium battery, and the intrinsic modal component (IMF) was obtained and reconstructed. Secondly, the logistic mapping and adaptive learning rate were integrated into the hippopotamus optimization (HO) algorithm to avoid the iterative process falling into the local optimum. And the improved hippopotamus algorithm was used to optimize the network parameters of long short-term memory (LSTM), then an improved HO-LSTM model was established. Finally, based on the improved HO-LSTM model, the SOH prediction of lithium battery was carried out to improve the prediction accuracy. Based on the verification of lithium battery capacity data, the results indicate that compared with the single LSTM prediction model, the VMD-HO-LSTM model based on dual-strategy optimization improves the prediction accuracy by 49.6%~81.9%. Compared with the VMD-LSTM model, the battery prediction accuracy of the proposed model is improved by 23.4%~59.0%, and the prediction accuracy is 0.976~0.998. The established model and analysis method have better prediction effect on SOH of lithium battery. References | Related Articles | Metrics