Archive

31 August 2026, Volume 45 Issue 8 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

Steel shell-concrete matching relation of composite pylon section Liang Huanwei1, Xiao Fengjian2, Lu Yang2, Fu Lijun2, Li Pan2, Xu Chunrong3, Lin Yu3, Zhao Canhui1 2026, 45(8): 1-10.  DOI: 10.3969/j.issn.1674-0696.2026.08.01 Abstract ( )   PDF (5582KB) ( )   To optimize the design parameters of the steel shell-concrete composite pylon (SSCCP) section, an ideal matching relationship between steel shell and concrete was proposed for the composite pylon without longitudinal reinforcement, which was based on the principle of fully utilizing materials. The analytical expression of the ideal matching relation was derived through sectional strain distribution and axial force equilibrium. Considering the compound confinement effect provided by stiffeners and hook bars, a confinement model for the core concrete of the pylon was established, and the design strain of concrete was defined as the constrained peak strain. Parametric analysis reveals that the ideal steel ratio increases linearly with concrete strength, confinement effect, and axial compression ratio. When the axial compression ratio increases from 0.05 to 0.25, the required minimum confinement strength ratio rises from 1.00 to 1.50. Based on this matching relation, formulas for calculating the yield moment and ultimate moment of the section were derived. Compared with finite element results, the error for the ultimate moment was within 2%, and the error for the yield moment was within 8%. It is recommended in the design to achieve material matching under a high steel ratio by enhancing the concrete strength or strengthening the confinement effect, and to keep the axial compression ratio within 0.20 to ensure the ductility of the section. References | Related Articles | Metrics

Select

BIM design software development of railway continuous girder bridge deck system based on independent graphics engine Qi Chenglong 2026, 45(8): 11-21.  DOI: 10.3969/j.issn.1674-0696.2026.08.02 Abstract ( )   PDF (4742KB) ( )   To overcome the limitations of foreign BIM software in terms of geometric engine, system architecture and data security, a roadmap developed based on the domestic BIMBase platform was researched and determined. By conducting in-depth API development of graphics engine using C++, a set of specialized tools for BIM forward design of railway continuous girder bridge deck system were constructed, with the core architecture centered around three layers. Firstly, at the data layer, the geometric semantics and engineering logic of bridge deck system components were thoroughly analyzed. Based on this, a specialized parameterized data structure was defined. This structure was deeply adapted to the component management mechanism of domestic graphics engine. The main bridge deck system class was derived from the platform’s fundamental graphic object (BPGraphicElement), forming core container for design information. Secondly, at the component functionality layer, core capabilities such as complete 3D modeling, attribute dynamic management and geometric transformations (e.g., moving, copying) for various bridge deck system parts were developed by systematically overriding key virtual functions of the base classes. Finally, at the interaction and application layer, the software could intelligently calculate the longitudinal layout range of different bridge deck cross-sections along the route and automatically generate geometric sweep paths. By parsing predefined local XML configuration files, cross-sectional property values can be automatically assigned to each component with one click, thus driving the parameterized generation of the entire model, and rapid conversion from design intent to 3D results was achieved. Testing indicates that the software development achievements have effectively achieved the goal of domestic substitution, and its parametric modeling efficiency has been significantly improved compared to international mainstream platforms. Meanwhile, this study has accumulated crucial experience and validated the feasibility of the technical pathway for developing high-performance specialized BIM tools based on BIMBase and C++. References | Related Articles | Metrics

Select

Seismic performance and load-carrying capacity model of GFRP-steel composite tube confined columns Li Yunyu1, Zhang Xiaolong1, Zhang Hepeng2, Liu Jing1 2026, 45(8): 22-29.  DOI: 10.3969/j.issn.1674-0696.2026.08.03 Abstract ( )   PDF (3270KB) ( )   To address the engineering requirements for seismic reinforcement of bridge piers and columns, the seismic performance of circular reinforced concrete columns confined by GFRP (glass fiber reinforced polymer) and steel composite pipes was studied. One unconstrained and two composite tube conserained reinforced concrete columns with different GFRP layers (5 and 7 layers) were designed and fabricated, and quasi-static tests were conducted to systematically analyze their hysteresis curves, skeleton curves and energy dissipation capabilities. Based on the experiment results, finite element models (FEM) were developed through OpenSees, to study the influence rule of GFRP layer number and axial compression ratio on the ultimate bearing capacity of peak load. A peak load prediction model applicable to this composite pipe constrained reinforced concrete column was proposed. The research results indicate that the GFRP-steel composite confinement system significantly enhances the load-carrying capacity, ductility, and energy dissipation performance of the confined columns. The peak load positively correlates with the number of GFRP layers, but it exhibits relatively low sensitivity to variations in the number of layers. Furthermore, increasing the axial compression ratio can significantly enhance the peak load of the specimen, but the enhancement effect gradually weakens with the increase of the axial compression ratio. The calculated values of the established peak load prediction model have an error within ± 10% compared to the finite element results, indicating good engineering applicability. References | Related Articles | Metrics

Select

Structural mechanical performance of composite wet joints in large-span open-cut tunnels Lin Zhi, Shang Wu, Xu Zhanglong, Guo Xiande, Zhang Guojiao 2026, 45(8): 30-42.  DOI: 10.3969/j.issn.1674-0696.2026.08.04 Abstract ( )   PDF (14997KB) ( )   To address the safety issues associated with large-span prefabricated structures for open-cut tunnels under deep overburden soil conditions, a novel composite wet joint system was proposed, aiming to simplify on-site construction processes, enhance tunnel lining quality and reduce project costs. Taking a two-way eight lane open-cut tunnel on Chongqing Xinsen Avenue as the background, a three-dimensional solid-contact model was developed to perform mechanical analysis under progressive backfilling conditions. Additionally, a full-scale bending test was designed to investigate the failure mechanism of the composite joint under high axial compression. Numerical simulation results indicate that the arch top settlement of the structure measures 17.3 mm, with a maximum tensile stress of 1.92 MPa and a compressive stress of 12.30 MPa, all satisfying design requirements. The internal force analysis of the arch section during the entire process of assembly and backfilling reveals a maximum bending moment of 2 136 kN·m, and the minimum safety coefficient of the composite joint is 5.2, indicating sufficient safety reserve. Experimental research indicates that the composite joint exhibits a typical failure mode characterized by eccentrically loaded compression, and the increase of axial force can significantly enhance the ultimate bending moment and suppress crack propagation, while the influence of steel reinforcement overlapping length is relatively weak. These research findings provide a technical pathway that balances safety, economy and efficiency for the design of ultra-large-span prefabricated open-cut tunnels. References | Related Articles | Metrics

Select

Strength formation mechanism of vegetation ash and lime-fly ash stabilized coal gangue mixture Yu Baoyang1, Huang Yutian2, Miao Fei1, Wang Weiguo2 2026, 45(8): 43-48.  DOI: 10.3969/j.issn.1674-0696.2026.08.05 Abstract ( )   PDF (3824KB) ( )   To address the environmental issues caused by coal gangue and vegetation ash, the vegetation ash and lime-fly ash stabilized coal gangue mixture was prepared. Multi-age mechanical property tests were conducted on the mixture, and the strength formation mechanism of the mixture was investigated through X-ray diffraction (XRD), scanning electron microscopy (SEM), and molecular dynamics simulations of the generated products. The mechanical properties of the mixture exhibited an initial increase followed by a decrease as the vegetation ash content increased, with the optimal dosage of 30%. The dosage of vegetation ash improved the compactness of the aggregate-binder interfacial transition zone and reduced the Ca/Si ratio of the gel within the mixture, thereby enhancing the mechanical properties of both the gel and the mixture. Ultimately, incorporating an appropriate amount of vegetation ash into the lime-fly ash stabilized coal gangue mixture can improve its compactness and gel strength, which in turn increases the overall mechanical strength of the lime-fly ash stabilized coal gangue mixture. References | Related Articles | Metrics

Select

Mix design and performance of solidifier for multi-solid waste silty soil Zhang Chengyi1, Shu Bin1, He Zhiliang2, Wu Feihu2 2026, 45(8): 49-57.  DOI: 10.3969/j.issn.1674-0696.2026.08.06 Abstract ( )   PDF (3801KB) ( )   Addressing the challenges of treating silty foundations in traffic engineering construction, a solidifier for silty soil was developed by use of industrial solid wastes such as lithium slag, calcium carbide slag and slag. Based on the mixture design method, a mathematical model was established for the unconfined compressive strength (qu) of solidified soil and the composition of raw materials. The optimal mixing ratio (lithium slag: calcium carbide slag: slag = 0.282: 0.170: 0.548) was obtained through optimization, and the deviation between the predicted and measured qu of solidified soil under this ratio was only 1.09%. The research results show that the qu, splitting strength (fst), shear strength (τ), CBR value, and resistance to wet dry cycles of the improved dredged silt are significantly improved with increasing dosage of solidifier. Microscopic mechanism analysis (SEM, XRD, FT-IR) confirms that the solidifier initiates hydration reactions, generating substantial amounts of calcium silicate hydrate and ettringite. These hydration products continuously develop with age, effectively cementing the soil and significantly improving the engineering properties of silt, which makes it have good road performance. References | Related Articles | Metrics

Select

Compressive-torsional behavior of geopolymer concrete columns based on the axial compressive ratio Peng Cheng, Du Yunxing, Shi Caijun 2026, 45(8): 58-66.  DOI: 10.3969/j.issn.1674-0696.2026.08.07 Abstract ( )   PDF (5630KB) ( )   To investigate the effect of axial compression ratio on the compression-torsion performance of geopolymer concrete rectangular columns, one Portland cement concrete (PC) contrastive specimen and five geopolymer concrete (GC) specimens were prepared and subjected to monotonic torsion tests under constant axial compression. And combined with the standard limit values, the failure mode, torque-torsion angle relationship, ductility degradation and damage evolution laws were analyzed. The results show that GC and PC columns exhibit similar failure modes under combined compression and torsion. But the crack distribution of GC specimens is more uniform and finer, and the pre-peak torsional deformation is more complete. The peak torque of GC-0.3 is 52.20 kN·m, 2.7% higher than that of PC-0.3 at 50.82 kN·m, indicating favorable load-carrying capacity. As the axial compression ratio increases from 0.1 to 0.6, the cracking torque and peak torque of GC columns increase by 101.7% and 37.6%, respectively. When the axial compression ratio further increases to 0.65, the peak torque decreases to 59.17 kN·m, indicating an upper limit to the enhancement effect of axial compression on torsional bearing capacity. Increasing the axial compression ratio will accelerate post-peak damage accumulation and shift the failure mode from ductile torsion-bending failure to brittle shear-torsion failure. The ductility coefficient of GC-0.5 specimen is 16.3% lower than that of GC-0.1, indicating that high axial compression will accelerate the concentration of post-peak compression shear damage and weaken the deformation capacity. Based on the test results, it is recommended that, when GB/T 50010—2010 is applied to the torsional design of GC columns under axial compression, the upper limit for axial compression ratio enhancement should be increased from n=0.3 to n=0.5. After this modification, the mean predicted-to-tested ratio is 0.96, with a variation coefficient of 7.65%. References | Related Articles | Metrics

Traffic & Transportation+Artificial Intelligence

Select

Lane-changing risk in freeway off-ramp areas based on EF-GCN Fang Mingshan1, Li Yi2, Yu Shanchan3, Wu You2, Dong Xin2 2026, 45(8): 67-74.  DOI: 10.3969/j.issn.1674-0696.2026.08.08 Abstract ( )   PDF (3765KB) ( )   Accurately predicting vehicle lane-changing risk in freeway off-ramp areas is crucial for enhancing road traffic safety, as this risk exhibits characteristics of “short cycles, strong fluctuations, and rapid decay”. Addressing issues of traditional methods being difficult to quantify multiple vehicle interactions, single energy field easily ignoring spatial cascading effects, and graph convolutional networks (GCN) lacking physical meaning in graph structure weights and unable to adapt to dynamic traffic scenarios, a vehicle risk prediction model (EF-GCN) that combined energy field (EF) and GCN was proposed to optimize the accuracy of risk prediction. Firstly, an energy field model was constructed based on UAV aerial data, generating a global risk heat map that quantified physical interaction intensity. Secondly, the EF-GCN model was proposed to map risk values to dynamic graph weights, accurately capturing global risk cascading effects triggered by lane changes through neighborhood aggregation. Finally, GRU was introduced to establish a temporal prediction branch, which simplified the dual gate control structure to eliminate redundant historical information and capture the temporal dependencies of risk evolution. Experimental verification based on the Shanghai Central Ring Road dataset demonstrates that the EF-GCN model, benefiting from the fusion of energy field physical representation and GCN spatial features, significantly outperforms traditional dynamics and single energy field models in risk assessment accuracy. Compared to the EF-GCN-LSTM model, the EF-GCN-GRU reduces ERMS by 9.38% and EMA by 9.30%, while improving EMS performance by 17.89%. The significant advantage of the proposed model lies in the construction of a “physical mechanism support and spatiotemporal deep fusion” architecture, which maps risk heat maps to dynamic graph structures, endows GCN with physical meaning, and breaks through the bottleneck of static graphs. Simultaneously, it accurately captures global risk cascading effects triggered by lane changes, clearly revealing the physical evolution law of risk, that is, “accumulation-peak-decline”. References | Related Articles | Metrics

Select

Analysis and prediction of vehicle speed on exit ramp of tunnel-adjacent interchange section Luo Shuang1, Wang Yan1, Yang Shuqun2 2026, 45(8): 75-83.  DOI: 10.3969/j.issn.1674-0696.2026.08.09 Abstract ( )   PDF (4527KB) ( )   To study the driving speed characteristics of vehicles exiting the ramp in the small clear distance section of tunnel-adjacent interchange, four groups of tunnel-interchange scenarios with different clear distances on the Baomao Expressway were selected. Real vehicle driving tests were carried out, and the speed data of 41 drivers were collected through the Mobileye E630 device. The speed variation law, speed dispersion characteristics and speed distribution of characteristic cross-sections of the small clear distance sections adjacent to the interchanges in the tunnel were analyzed. The speed prediction models for vehicles at the diversion points and small nose points were established and verified. The results show that the vehicle speed on the section with a small clear distance shows a significant downward trend, and the deceleration amplitude increases as the clear distance decreases. Small clear distance road sections belong to a strong coupling environment between vehicles and roads, with strong driving behavior convergence. Drivers’ deceleration behavior is more severe on small clear distance road sections than in conventional tunnels and interchanges, especially when the clear distance does not exceed 50m, the deceleration amplitude can reach 23.7%. The speed at each characteristic section of the small clear distance road section with a clear distance not exceeding 170 meters is significantly lower than that of the conventional section. The average absolute percentage errors of the prediction model are all less than 10%, demonstrating a relatively high prediction accuracy. References | Related Articles | Metrics

Select

Construction risk assessment of railway subgrade in dangerous mountainous areas based on GRA-TOPSIS model Yan Linjun1,2, Lin Jianen1, Zhang Jingwei1,2 2026, 45(8): 84-94.  DOI: 10.3969/j.issn.1674-0696.2026.08.10 Abstract ( )   PDF (1985KB) ( )   To objectively evaluate the construction risks of railway subgrade in perilous mountainous areas, an improved GRA-TOPSIS comprehensive evaluation model was established in view of the complexity and fuzziness of the risk assessment of subgrade construction. Firstly, the WBS-RBS method was adopted to construct a coupling matrix for identifying risk factors, and a comprehensive evaluation index system for railway subgrade construction risks in perilous areas was established from five aspects such as human, material, technology, management and environment. Secondly, the game theory was used to combine the order relation analysis method and the CRITIC method for weighting, and the comprehensive weights of the indicators were determined. Finally, according to case studies and data analysis, the GRA-TOPSIS comprehensive evaluation model was constructed to conduct a specific assessment of construction risks for a total of 5 subgrade sections within the case interval. The research results show that the subgrade sections K2 and K4 perform well overall, followed by K1, with medium construction risks. In contrast, the subgrade section K3 has a higher risk, and K5 has the highest risk, both with high construction risks. The comprehensive risk evaluation grade of the subgrade is “medium to high risk”, verifying the feasibility of the proposed evaluation model. Specific risk control measures are proposed for the risk evaluation results of each subgrade section in the case, providing a new idea for the risk evaluation of railway subgrade construction in perilous and complex areas. References | Related Articles | Metrics

Select

Control strategy and emission optimization of CAV on freeway in foggy days Wang Jun1, Luo Zhongbin2,3, Fu Lijia3, Yang Xun3 2026, 45(8): 95-107.  DOI: 10.3969/j.issn.1674-0696.2026.08.11 Abstract ( )   PDF (8929KB) ( )   To address the issue of increased energy consumption and emissions caused by frequent acceleration and deceleration due to low visibility on highways in foggy days, a dynamic-weight car-following control strategy for connected and automated vehicles (CAVs) based on model predictive control (MPC) was proposed, and its asymptotic stability was proved theoretically. Driving simulation experiments on highway sections in actual foggy days were conducted. Based on the driving simulation data, a car-following model for human-driven vehicles (HDVs) under foggy conditions was calibrated, and the accuracy of the calibration results was verified. Finally, comparative experiments on the fuel consumption and emission characteristics of mixed traffic flow in fog days were carried out based on the SUMO simulation platform. The research results show that the proposed dynamic-weight CAV control strategy effectively achieves the goals of energy saving and emission reduction, significantly narrowing the disparities in fuel consumption and emissions among vehicles. Moreover, its energy-saving and emission-reduction performance is superior to those of existing strategies such as the conventional fixed-weight MPC and eco-cooperative adaptive cruise control. The findings can provide a theoretical and methodological basis for the management and control of CAVs on highways in foggy days. References | Related Articles | Metrics

Modern Traffic Equipment

Select

Research on local obstacle avoidance planning and control of intelligent vehicle Niu Limin1,2, Wen Zeyun1, Cao Jingjie1, Yang Yingwei1, Wang Menghan1 2026, 45(8): 108-117.  DOI: 10.3969/j.issn.1674-0696.2026.08.12 Abstract ( )   PDF (9145KB) ( )   Addressing the issues of rough paths and poor tracking performance in local obstacle avoidance planning for vehicles, a that combined dynamic programming (DP) with quadratic programming (QP) local, was proposed and coupled with model predictive controller (MPC) model for prediction. Firstly, a preliminary path satisfying obstacle avoidance constraints was generated by use of dynamic programming algorithms. Subsequently, to address the issues of poor smoothness and insufficient safety distance in this preliminary path, quadratic programming was introduced for optimization and smoothing. Finally, the optimized path was tracked and controlled by a coupled lateral and longitudinal MPC control algorithm. The results show that, in the process of avoiding obstacles, the path optimized by DP+QP is smoother and more comfortable compared to the original path generated by DP. Meanwhile, the coupled lateral and longitudinal control strategy of the MPC algorithm can effectively reduce control errors, enabling the vehicle to travel smoothly in complex environments. References | Related Articles | Metrics

Select

Road slope identification method jointly driven by kinematic Kalman filtering and CNN-LSTM Deng Chaoyang1, Li Jianlu1, Huang Zhaoyong1, Cai Minrui2, Zheng Yuan2, He Guolin2 2026, 45(8): 118-125.  DOI: 10.3969/j.issn.1674-0696.2026.08.13 Abstract ( )   PDF (4525KB) ( )   With the rapid development of the new energy vehicle industry, accurate road slope estimation is of great significance for optimizing vehicle control strategies and improving driving safety and economy. Traditional kinematic Kalman filtering methods suffered from large estimation errors in sections with severe acceleration variations during braking, owing to the neglect of vehicle pitch angle effects. To address this issue, a joint-driven approach combining kinematic Kalman filtering and deep learning was proposed for intelligent road slope estimation. In the proposed method, Kalman filtering was employed to process normal driving conditions, and the slope estimation values of sudden acceleration section during braking conditions were dynamically corrected, utilizing trained convolutional neural network (CNN) and long short term memory (LSTM) network models. Real vehicle road test results demonstrate that the proposed method yields significantly lower mean absolute error compared to the single Kalman filtering method in typical driving segments, confirming its effectiveness and robustness under high dynamic conditions. References | Related Articles | Metrics

Select

Simulation analysis of axial impact failure of CFRTP hat-shaped thin-walled beam Xu Xianzhe1, Jia Zengrui2, Jiang Yuanshun1, Guo Zhenxi1 2026, 45(8): 126-134.  DOI: 10.3969/j.issn.1674-0696.2026.08.14 Abstract ( )   PDF (6613KB) ( )   On the premise of ensuring the mechanical properties such as structural strength and stiffness of the vehicle body, the use of composite materials to partially or completely replace metal materials is an effective means of achieving vehicle lightweight. Front longitudinal beam (FLB), located beneath the engine compartment, plays role in energy absorption and occupants protection during frontal crashing scenarios. Aim to study the dynamic mechanical properties of CFRTP hat-shaped structure FLB of vehicle, the dynamic mechanical performance parameters of CFRTP were measured, by designing and manufacturing composite material dynamic mechanical performance test fixtures. A finite element simulation model of axial compression impact of hat-shaped beam was established in Hypermesh and solved in LS-DYNA program. Mechanical properties and energy absorption performance of hat-shaped beam were studied by changing the cross-sectional shape of the hat-shaped beam (single hat-shaped, double hat-shaped) and the fiber layup angle (［± 45°］, ［0°/90°］). The research shows that under the same layup condition, the loading stiffness of double hat-shaped beams is greater than that of single hat-shaped beam, and under the same cross-sectional shape condition, the stiffness of ［± 45°］ layup is greater than that of ［0°/90°］ layup. The peak force of double hat-shaped beam is 13.3% and 14.3% higher than that of single hat-shaped beam in ［±45°］ and ［0°/90°］ layup, respectively, and the cross-sectional shape has a significant influence on the peak force. The single hat-shaped beam ［0°/90°］ layup has a 12.0% increase in SEA compared to the ［± 45°］ layup. The double hat-shaped beam ［0°/90°］ layup has an 11.0% increase in SEA compared to the ［± 45°］ layer. The change in layering method has a significant impact on SEA of the structure. References | Related Articles | Metrics