Current Issue

2026, Volume 26,  Issue 6

column
  Previous Issue
Display Method:
The Others
Cover and Contents of Vol.26, No.6, 2026
Cover Contents
2026, 26(6): .
FullText HTML (42) PDF (285KB) (22)
Special Column on New Materials and Structural Systems for Bridges
Technological evolution and span breakthrough of continuous rigid-frame bridges
WU Ming-yuan, LIU Yong-jian
Abstract: More> To systematically analyze the technical logic behind span breakthrough in continuous rigid-frame bridges (CRFBs), address industry challenges such as long-term deflection and cracking of girders under long-span conditions, and improve the design method of such bridges, engineering cases from multiple typical CRFBs were compiled. With span breakthroughs as the core thread, the theoretical formula derivations were integrated with finite element simulations to analyze the evolution patterns, mechanical mechanisms, and engineering adaptability of key technologies across various stages. The practical value of the hybrid girder continuous rigid-frame system, inclined web steel-concrete connection section, and double-cantilever construction technology was verified based on practical projects including the Tao'er River Grand Bridge and Guijiang Grand Bridge. Analysis results reveal that, the technical development of CRFBs can be categorized into three stages: Foundational theory refinement, lightweight optimization, and double-cantilever construction advancement. For prestressed concrete CRFBs, the reasonable upper limit of the span is constrained to approximately 300 m due to long-term mid-span deflection and girder cracking. Early hybrid girder CRFBs are restricted by the full-span steel girder hoisting method and the design principle of locating the connection section at one-third of the main span, resulting in an ultimate span limit of approximately 375 m. The inclined web steel-concrete connection section significantly enhances force transfer performance in high-stress zones and offers flexible placement along the full span. When combined with cantilever assembly of steel girders, this structural configuration theoretically enables a span capacity of 500 m class. Furthermore, the CRFB scheme demonstrates superior comprehensive competitiveness within the 200 - 400 m span range. This study elucidates the inherent technical logic governing span breakthrough in CRFBs, providing a theoretical reference and practical support for design optimization, engineering applications, and the future technical breakthrough toward 500-meter-class CRFBs.
2026, 26(6): 1-20. doi: 10.19818/j.cnki.1671-1637.2026.313
FullText HTML (69) PDF (11955KB) (49)
Longitudinal movement characteristics and rational restraint systems of long-span railway suspension bridges under train load
WANG Hui, SHEN Rui-li, LIU Gao, XIN Hao-hui, YIN Ru-yang
Abstract: More> To investigate the longitudinal movement characteristics of long-span railway suspension bridges under train loads and determine a rational longitudinal restraint system, a field test was conducted to analyze the longitudinal movement responses at the girder ends during train passage. A refined longitudinal dynamic analysis model was developed based on an in-service railway suspension bridge, and its reliability was verified by comparing with measured results. Based on this, analysis models of different longitudinal restraint systems were established, and a transient dynamic analysis method was adopted to compare the longitudinal vibration responses at the girder ends and the mechanical behaviors of key connecting components under train loads. The results indicate that when a single-tower fixed restraint is adopted, the peak longitudinal displacement of the movable bearing caused by train passage is 9.80 mm. Due to the influence of bearing friction and girder-rail interaction, the bearing fails to reset rapidly after the train passes. The developed longitudinal dynamic model accurately characterizes the dynamic variations of longitudinal bearing displacement during the whole process of train passage, and the peak displacement agrees well with the test results. Increasing longitudinal restraint stiffness and adding supplemental damping can both effectively suppress the longitudinal vibration of the stiffening girder. Under the floating system, the longitudinal displacement amplitude at the girder end reaches 79.2 mm, which decreases to 43.1 mm after the central buckle is set and further decreases to 30.5 mm after the damper is added. The central buckle significantly improves the mechanical behaviors of short hangers. Without the central buckle, the maximum bending stress of the hanger is 23.57 MPa, which decreases to 8.03 MPa after installation, with a reduction of 65.9%. The stress amplitude of the central buckle shows an increasing trend with the increase of longitudinal restraint stiffness and reaches 191.59 MPa and 205.92 MPa after adding dampers or considering bearing friction, respectively. Excessive longitudinal restraint stiffness doubles the bending stress of short hangers, causes stress concentration in the central buckle, and seriously threatens the service life of mid-span connecting components. The design of longitudinal restraint systems for long-span railway suspension bridges should comprehensively consider the longitudinal displacement control at the girder ends and the mechanical balance of short hangers to avoid unfavorable internal force concentration of components caused by excessive longitudinal restraint stiffness. The research results can provide a reference for the design optimization and healthy operation and maintenance of longitudinal restraint systems for long-span railway suspension bridges.
2026, 26(6): 21-35. doi: 10.19818/j.cnki.1671-1637.2026.232
FullText HTML (36) PDF (17792KB) (25)
Axial compressive performance of round-ended UHPC-filled aluminum alloy tube columns after lateral impact
ZHANG Wei-wei, XING Zhi-quan, CHEN Yu, ZHAO Yan-gang, YOSHIYAMA Hiroshi, LIN Si-qi, SONG Tian-yi, GHAFORY-ASHTIANY Mohsen, KUMAR Manish
Abstract: More> To investigate the influence mechanism of lateral impact damage on the axial compressive performance of round-ended UHPC-filled aluminum alloy tube (RE-UCFAT) columns, drop-weight impact tests were first conducted to apply initial damage, and axial compression tests were then carried out to systematically quantify the weakening effects of impact damage on the residual ultimate load-carrying capacity and axial stiffness of members. A refined parameter-calibrated finite element model was established to reveal the evolution mechanism of impact damage. Based on the unified theory and the idea of effective confinement zoning, a calculation method for the ultimate load-carrying capacity applicable to RE-UCFAT columns was proposed. By decoupling the cross-sectional solid damage and the overall stability reduction effect, a prediction model for residual load-carrying capacity containing energy-axial compression cross-coupling was established. Research results indicate that the round-ended aluminum alloy tube has a good confinement effect on the core UHPC, and the typical failure mode of the member is weak-axis shear failure. The lateral impact leads to significant degradation of the load-carrying capacity and stiffness of the member, and the performance attenuation caused by straight-edge damage is significantly greater than that caused by curved-edge damage. Under a high axial compression ratio, the impact damage induces a transition in the failure mode. The straight-edge damage causes the member to shift to combined bending-shear failure by introducing the second-order effect, while the curved-edge damage induces strong-axis shear failure by introducing punching shear damage. Validation results show that the established numerical model and the two types of analytical models for load-carrying capacity have high accuracy and can provide a theoretical basis for the impact-resistant design and assessment of this novel type of composite columns.
2026, 26(6): 36-51. doi: 10.19818/j.cnki.1671-1637.2026.317
FullText HTML (34) PDF (8393KB) (21)
Mechanical performance of UHPC keyed epoxy joints under combined bending and shear action
PAN Ren-sheng, ZOU Ze-peng, ZHOU Xuan, GONG Zhen-min, PENG Jian-xin
Abstract: More> To investigate the mechanical performance of keyed epoxy joints in ultra-high-performance concrete (UHPC) under combined bending and shear action, shear tests on 14 UHPC keyed epoxy joints were conducted with parameters including the bending-shear ratio (a/h), lateral stress, and steel fiber content (0-2%). The failure modes, interface relative sliding, and shear strength variation laws of the specimens under combined bending and shear action were studied. According to the results, the shear strength of UHPC keyed epoxy joints decreases with the higher bending-shear ratio on the whole. However, there is an obvious boundary bending-shear ratio for the influence of the bending-shear ratio on the shear strength. When a/h≤0.4, the influence of the bending-shear ratio on the joint shear strength is small. When a/h>0.4, the influence of the higher bending-shear ratio is relatively significant. When the UHPC steel fiber content increases from 0 to 2%, the shear strength of the combined bending and shear joint specimens rises by 50.3%. The shear strength of both direct shear and bending-shear joint specimens exhibits a linear growth trend with the increase of lateral stress. When the bending-shear ratio is relatively large, the higher lateral stress significantly enhances the shear strength of the joint, indicating an interactive effect between lateral stress and the bending-shear ratio on the joint shear strength. Based on the experimental data, a shear capacity calculation formula for the UHPC keyed epoxy joint was proposed considering the bending moment. Verification results demonstrate a good agreement between the calculated results and the experimental data.
2026, 26(6): 52-62. doi: 10.19818/j.cnki.1671-1637.2026.235
FullText HTML (42) PDF (24473KB) (18)
Fatigue performance of hot-rolled thickened U rib-to-deck double-sided welds based on structural stress method
LIU Hong-xi, LIU Yu-qing, CHANG Zhi-jun, CHEN Hu-cheng, XU Xiao-qing
Abstract: More> Finite element analysis was conducted on orthotropic steel deck models incorporating U ribs of different dimensions to elucidate the fatigue performance of hot-rolled thickened U rib-to-deck double-sided weld details and investigate the influence of increased thickness at U rib limb ends on weld fatigue strength. Based on the structural stress method, the influence of U rib dimensions on the fatigue performance of U rib-to-deck welds was investigated. Furthermore, the fatigue life of the constant thickness U rib and hot-rolled thickened U rib-to-deck weld details was analyzed according to the master S-N curve. The results indicate that the equivalent structural stress at the inner weld of the U rib-to-deck joint is significantly higher than that at the outer weld, thus making the inner weld toe more prone to fatigue crack initiation. The adoption of 8-12 mm hot-rolled thickened U ribs effectively reduces the equivalent structural stress at the inner weld of the U rib-deck joint by 12.9%. Based on the master S-N curve with a 95% confidence interval, compared to constant thickness 8 mm U ribs, the fatigue life of the inner side weld details by employing 8-12 mm and 12-16 mm hot-rolled thickened U ribs increases by 54.2% and 174.8% respectively. Hot-rolled thickened U ribs improve the fatigue performance of U rib-to-deck weld details by locally increasing the thickness at the limb ends. The combination of thickened limb-end configurations and double-sided welds provides a viable approach for improving the fatigue performance of orthotropic steel decks.
2026, 26(6): 63-71. doi: 10.19818/j.cnki.1671-1637.2026.234
FullText HTML (47) PDF (5129KB) (24)
Repair of damaged hinge joints in hollow slab bridges using local prestress from U-shaped Fe-SMA rebars
DONG Zhi-qiang, ZOU Cui, SUN Xin-liang, ZHU Hong
Abstract: More> To investigate the repair effect of local prestress generated by iron-based shape memory alloy (Fe-SMA) rebars on damaged hinge joints in hollow slab bridges, a rapid repair method for damaged hinge joints based on the local prestress technique using Fe-SMA rebars was proposed. In this method, the conventional U-shaped rebars originally arranged in the hinge joints were replaced with U-shaped Fe-SMA rebars, and the local transverse prestress generated after thermal activation was utilized to rapidly restore the transverse load transfer performance of the hinge joints. A 1/4-scale model consisting of six hollow slab beams was constructed, damage was artificially introduced into the hinge joints, and static loading tests were conducted under multiple loading conditions. The proposed method was then used to repair the damaged hinge joints, and the above loading tests were repeated. An ultimate loading test was finally carried out on slab beam No. 4. Finite element analysis was also conducted to evaluate the repair effect under different levels of hinge joint damage and Fe-SMA recovery stress. The experimental and finite element analysis results show that hinge joint damage significantly weakens the load transfer performance, leading to irregular fluctuations in the load transverse distribution coefficient under various loading conditions. After repair, the load transverse distribution tends to be uniform. The midspan deflection of the loaded slab beams decreases by 23.2%-33.7%, the peak value of the load transverse distribution coefficient decreases by 14.29%-27.17%, and the longitudinal reinforcement strain decreases by 25.6%-34.0%. As the damage level increases, the load transfer capacity of the hinge joints decreases, and their condition transitions from intact to damaged. After applying a local prestress of 330 MPa, the load transfer capacity of the hinge joints under various damage levels is essentially restored to the intact level. Increasing the recovery stress of the Fe-SMA rebars enhances the load transfer performance. However, excessively high prestress increases the interfacial stress level, and the enhancement effect gradually approaches saturation.
2026, 26(6): 72-89. doi: 10.19818/j.cnki.1671-1637.2026.127
FullText HTML (46) PDF (25539KB) (21)
Composite box girder beam element with corrugated steel webs considering shear force interaction in flange plates
LI Xia-yuan, ZHOU Man, LI Li-feng, CHEN Jun, FU Li-xiang, KANG Ai-hong
Abstract: More> To investigate the shear force distribution in composite box girder with corrugated steel webs (CSWs), a novel box girder beam element (TBTF-CSW) comprehensively was proposed considering the shear force distribution relationship between CSWs and flange plates, as well as the influence of diaphragm constraints. By introducing shear strain transfer coefficients for the flange plates, the shear forces carried by CSWs, top and bottom flange plates were effectively decoupled. Considering the variable cross-section effect, analytical expressions for the shear force distribution in variable cross-section composite box girders with CSWs were derived, and a theoretical analysis model was established considering shear deformation influence in both the flange plates and CSWs. Based on the energy variational principle, the controlled differential equations that consider the shear force interaction of the flange plates were formulated. The shear deformation of the CSWs was then introduced as an additional degree of freedom. The homogeneous solutions of the controlled differential equations were used to construct interpolating functions for the generalized displacements, including vertical deflection, CSWs equivalent shear strain, and equivalent bending rotation. By integrating the finite element method, a two-node six-degree-of-freedom box-girder beam element considering the shear deformation influence of both CSWs and flange plates was developed. A series of typical numerical examples were analyzed to verify the accuracy and applicability of the TBTF-CSW beam element in calculating the shear stress and shear force distribution for composite box girders with constant and variable cross-sections. The results indicate that in the constant cross-section composite box girders with CSWs, the CSWs carry the majority of the shear force, with the distribution primarily affected by material characteristics and geometric parameters. The diaphragms and concentrated loads introduce local disturbances. In variable cross-section composite box girders, the axial force in the bottom flange plate leads to significant variable cross-section effect. Either the bottom flange plate or CSWs may experience "shear overdistribution". In such cases, it becomes unsafe to assume that the CSWs carry the entire shear force. The proposed TBTF-CSW beam element provides an effective support for theoretical analysis of shear force distribution and engineering design for composite box girders with CSWs.
2026, 26(6): 90-103. doi: 10.19818/j.cnki.1671-1637.2026.193
FullText HTML (25) PDF (5533KB) (6)
Experimental study on high-temperature creep-fatigue sequential coupling performance of Q345 steel for bridge
XIN Hao-hui, HUANG Xu, ZHAO Qian-yu, LIU Gao, NI Ya, YIN Ru-yang
Abstract: More> To meet the practical demand for fatigue performance assessment of bridge steel structures after fire, Q345 low-carbon alloy steel, which is widely used in bridge engineering, was selected as the research object. A total of 15 standard cylindrical dog-bone specimens were designed in 4 categories and 7 groups. The influence law of different creep damage durations under sustained loading on the residual deformation and fatigue performance of the steel after cooling was systematically investigated. The results show that as the high-temperature creep time increases from 0 to 0.7 times the creep rupture time, the average fatigue life of the sequential creep-fatigue coupling specimens decreases from 3.22×105 cycles to 1.15×105 cycles, representing a reduction of 64.4%. The sequential creep-fatigue coupling effect transforms the fracture morphology into a cup-and-cone shape and shrinks the fatigue propagation zone. This is mainly due to the reduction in material load-bearing capacity caused by oxidation decarburization and creep damage, which accelerates fatigue failure. Different high-temperature creep times have little effect on the dynamic elastic modulus of the specimens. The softening ratio of the elastic modulus is close to 1.0, and the initial dynamic elastic modulus decreases by only 8.7% at maximum. The critical cumulative plastic strain decreases significantly with increasing creep time, with a maximum reduction of 55.6%, while the plastic strain accumulation rate remains basically unchanged. As the creep time increases, the high-temperature creep damage reduces the cumulative plastic deformation tolerance of the sequential fatigue process, causing the material to accelerate softening and fail at lower plastic strain levels. This study can provide an experimental basis and theoretical reference for the fatigue performance assessment and residual life prediction of fire-damaged bridge steel structures.
2026, 26(6): 104-114. doi: 10.19818/j.cnki.1671-1637.2026.314
FullText HTML (28) PDF (5246KB) (7)
Creep property prediction of alkali-activated fly ash-slag concrete
HUANG Dun-wen, ZOU You-bao, DOU Shu-hao, XIA Li-peng, HE Jun, PENG Hui
Abstract: More> To discuss the creep prediction method for alkali-activated fly ash-slag concrete, creep tests on alkali-activated fly ash-slag concrete under varying stress-strength ratios, strength levels, and loading ages were carried out. The similarities and differences between alkali-activated concrete and cement concrete in the basic creep and drying creep were explored. Based on the existing creep data, a modified prediction method based on the creep model of the European Concrete Committee (CEB-FIP) was proposed. The results reveal that the critical point for linear creep in alkali-activated concrete occurs at a stress-strength ratio of 0.6-0.8. Additionally, the basic creep of alkali-activated concrete exceeds that of cement concrete with equivalent strength, while its drying creep develops more slowly during the early loading stage. The variation patterns of the creep coefficient with strength and loading age resemble those of cement concrete. The hyperbolic power function from the CEB-FIP creep model remains applicable for predicting the creep of alkali-activated concrete. According to the parameter analysis, the internal relative humidity term affecting the notional creep coefficient needs adjustment so as to match the higher basic creep of alkali-activated concrete. In addition, the coefficient describing the time development of creep should also be modified with respect to internal relative humidity. The established creep prediction model offers a viable approach for investigating the creep property of alkali-activated concrete structures.
2026, 26(6): 115-122. doi: 10.19818/j.cnki.1671-1637.2026.128
FullText HTML (25) PDF (1508KB) (9)
Traffic Information and Control
Infrastructure enhanced motion planning for automatic driving on unstructured roads: A case study of ETC stations
LEI Ming-yue, LAI Jin-tao, HU Jia
Abstract: More> To enhance the safety, generality, and practicality of the motion planning method for connected and automated vehicles based on optimal control, an infrastructure enhanced motion planning method was proposed for connected and automated vehicles. A motion control system based on a control barrier function was constructed to ensure strong constraints for safe collision avoidance under nonlinear system configuration, enabling vehicles to respond proactively and avoid over-the-horizon obstacles. A coordinate system with the optimized path as the reference line was proposed. It eliminated the dependence on road centerlines, and adapted to motion planning in unstructured road environments. An "infrastructure assisting vehicle" motion planning method was proposed, breaking the limitations of low computational efficiency, low safety, and poor practicality inherent in the traditional "infrastructure driving vehicle" motion planning method. Simulation testing was conducted on the proposed motion planning method, with electronic toll collection (ETC) stations taken as a typical case. The research results show that on typical unstructured road scenarios characterized by transient lane markings, missing lane markings, and lane-occupying construction, the proposed method is superior to traditional onboard motion planning models. The driving risk is reduced by 11.34%, and travel efficiency is increased by 17.36%. The average single-response runtime is less than 0.05 s for each module. The proposed motion planning method effectively reduces over-the-horizon risks, improves the efficiency of traversing unstructured roads, enhances real-time performance in algorithm application, and maintains robustness under varying vehicle-infrastructure communication conditions.
2026, 26(6): 123-136. doi: 10.19818/j.cnki.1671-1637.2026.078
FullText HTML (27) PDF (6955KB) (5)
Construction of VSSM-CNN detection network for nighttime road traffic accidents
YANG Yang, CHEN Xian-tian, WANG Jian-yu, PU Zi-yuan, ZHAO Hong-zhuan, YUAN Zhen-zhou
Abstract: More> To enhance the automatic detection performance of road traffic accidents in nighttime scenarios, a VSSM-CNN encoder-decoder architecture was constructed based on a visual state space model (VSSM) and a convolutional neural network (CNN), and an unsupervised traffic accident detection framework oriented to this scenario was proposed. By referencing the idea of feature fusion and based on existing visible-light images as appearance features, fine-grained optical flow information of video sequences was further extracted and processed using a recurrent all-pairs field transform (RAFT) optical flow estimation algorithm combined with a convolutional long short-term memory (ConvLSTM) module to represent traffic motion states. The two types of features, appearance and motion, were fused and input into a feature encoder with the VSSM as a backbone network for global feature extraction, and a CNN was adopted as a decoder architecture to recover images layer by layer to enhance local details, so as to strengthen feature extraction efficiency and utilization effects. A triple loss function combination of mean squared error, mean absolute error, and structural similarity was adopted, and weight proportions were determined through Bayesian optimization to improve the robustness of the model to abnormal structures and noise during nighttime image reconstruction. Research results indicate that the area under the receiver operating characteristic curve (ROC-AUC) and area under the precision-recall curve (PR-AUC) of the proposed method are 0.818 and 0.765, respectively, which improve by 22.6% and 20.3%, respectively, compared with traditional generative networks; among them, the ROC-AUC achieves the highest value in the comparison with multiple existing methods; the triple loss function shows the strongest model performance improvement ability in ablation experiments; the research method achieves the lowest model complexity and a favorable detection speed, fully demonstrating its stability in anomaly recognition and potential for deployment and application. The research results effectively improve the traffic accident detection ability in nighttime scenarios, can be further extended to traffic accident detection in low-illumination and low-visibility scenarios, and provide a feasible technical solution and reference for related applications.
2026, 26(6): 137-152. doi: 10.19818/j.cnki.1671-1637.2026.076
FullText HTML (24) PDF (9562KB) (17)
Adaptive offloading model for remote driving takeover task based on cloud-edge collaboration
ZHAO Hong-zhuan, WANG Yi-chen, ZHANG Ji-kang, YUAN Quan, WANG Jian-qiang, YANG Liang-yi, WANG Tao, ZHOU Dan
Abstract: More> An adaptive task offloading model based on cloud-edge collaboration was established, and the problems of high latency and unstable connection during remote driving takeover caused by network fluctuation and insufficient computing power were deeply analyzed. Three types of tasks, namely real-time control, computation-intensive, and interactive service tasks, were defined, and a two-level priority system of urgent and general levels was set to accurately distinguish the differentiated requirements of different tasks for latency and reliability. A collaborative computing environment integrating cloud center, edge node, and onboard terminal was constructed; a hierarchical offloading rule dynamically allocating computing nodes based on task priority and adaptively adjusting resource weights combined with real-time network bandwidth and edge load was proposed; a breakpoint resume mechanism based on backup nodes was studied to enhance the robustness of the system in unstable environments. A decision model with global minimum latency as the optimization objective was constructed using a dynamic programming algorithm, and a corresponding reward function was set to quantitatively evaluate the effectiveness of different offloading strategies. A dedicated dataset was constructed based on 11 parameters such as task data volume and processor frequency, and comparative experiments were designed to systematically study the performance of the model under dynamic load and different resource states. The research results indicate that under the scenario of dynamically changing edge load, the reward value obtained by the proposed adaptive offloading strategy is increased by 13.2% compared to the traditional fixed-threshold edge computing method; after introducing cloud collaborative computing, the overall reward value of the system is increased by 23.6% compared to the edge-only computing scheme; especially when the edge node load exceeds 60%, the proposed strategy can effectively reduce the task blocking rate by 45%.
2026, 26(6): 153-166. doi: 10.19818/j.cnki.1671-1637.2026.030
FullText HTML (33) PDF (4721KB) (9)
Road pothole segmentation network based on multi-scale feature fusion and channel feature adaptation
WANG Si-yu, FANG Hong-su, YANG Wei, ZHOU Yong-jun
Abstract: More> To accurately identify road potholes and the improve model generalization segmentation ability, a road pothole segmentation network (potholes-FBConvNet) based on astrous spatial pyramid fusion (ASPF) and feature adaptation was proposed. The convolutional neural network model (ConvNext) was used as the main structure for multi-scale feature extraction at different network depths. The unified perception and resolution network architecture (UPerNet) was used as the neck fusion network. The multi-level feature representations extracted through the main network were fully utilized, and the ASPF module was introduced into the neck network to further enhance its ability to capture multi-scale contextual feature information. A feature adaptation module was embedded between the neck and decoding head to perform adaptive calibration on the channel and spatial features before neck fusion feature decoding. In this way, the network segmentation performance was effectively improved. A total of 2 097 road pothole segmentation databases were separately collected and meticulously annotated (1 065 generally damaged potholes, 507 completely damaged potholes, 525 severely damaged potholes), and comparative and ablation experiments were conducted in sequence to verify network performance. The experimental results show that, the proposed potholes-FBConvNet model achieves an intersection over union (IoU), similarity coefficient, precision, and recall of 85.43%, 92.14%, 92.96%, and 91.33% on generally damaged potholes. On completely damaged potholes, it reaches 87.76%, 93.48%, 92.33%, and 94.67%. On severely damaged potholes, it reaches 90.76%, 95.15%, 95.41%, and 94.90%. Compared with 16 typical comparison models based on Transformer and Conv, the proposed segmentation network has the optimal generalization segmentation ability and robustness.
2026, 26(6): 167-185. doi: 10.19818/j.cnki.1671-1637.2026.031
FullText HTML (27) PDF (21730KB) (7)
Transportation Planning and Management
Forecasting ride-hailing demand via contextual spatiotemporal cross-attention mechanism
ZHAO Xia, SHI Zhuo-ya, LI Zhi-hong, LIU Jian-feng, WU Meng-lin, LI Chen-ji
Abstract: More> To accurately characterize passengers' temporal preferences for frequent travel within specific geographic areas, or their spatial preferences for frequently visited locations within given time windows, this paper proposed a ride-hailing demand prediction model named ST-BiAformer (Spatiotemporal Bidirectional Association Transformer), which integrates a contextual spatiotemporal cross-attention mechanism. The model captured sequential dependencies in travel demand by constructing a contextual temporal correlation module; designed a spatiotemporal cross-attention mechanism to cross-extract spatial (or temporal) dependencies within specific temporal (or spatial) contexts, thereby revealing recurrent travel demand patterns along targeted spatiotemporal dimensions; and further developed a spatiotemporal fusion module to enhance the model's performance in both single-step and multi-step demand forecasting. The proposed model was evaluated on multiple datasets through a series of comparative, ablation, and robustness experiments, assessing its predictive performance under various spatiotemporal scenarios. Experimental results demonstrate that, compared with the optimal baseline model, the mean absolute error and root mean square error are reduced by 7.21% and 5.54%, respectively, demonstrating improved overall prediction accuracy. In the 5 min, 10 min, and 15 min forecasting tasks, both error metrics decrease by 1% - 7%, indicating sound single-step and multi-step predictive capability. The model achieves the best performance only when all three constituent submodules work jointly, enabling comprehensive modeling of travel demand dependencies along both contextual temporal and spatiotemporal correlation dimensions. The proposed model is expected to precisely match users' ride-hailing demands across heterogeneous spatiotemporal scenarios at different time periods or locations, thereby providing technical support for enhancing supply-demand scheduling in urban ride-hailing services.
2026, 26(6): 186-197. doi: 10.19818/j.cnki.1671-1637.2026.075
FullText HTML (28) PDF (5317KB) (8)
Collaborative optimization model of fleet dynamic scheduling and supporting facility layout considering SAEV charging demand
HAN Fei, CAO Wan-biao, WANG Jian, LI Yan, SUN Chao
Abstract: More> To achieve the collaborative optimization of fleet scheduling and supporting facility layout of shared autonomous electric vehicle (SAEV), a multi-objective nonlinear programming model was established, with the objectives of minimizing the SAEV fleet size, total vehicle travel distance, total passenger travel time, and construction cost of charging and parking facilities. Based on a time-expanded network, the dynamic OD travel demand of passengers, the dynamic scheduling strategy of the SAEV fleet, and the spatiotemporal displacement of passenger flows were described. Furthermore, the capacity constraints of nodes and arcs in the network were utilized to characterize the congestion effects of charging and parking facilities in traffic zones and connecting roads, respectively. Distinguished from traditional models, several constraints were considered, including the charging demand of the SAEV fleet, dynamic conservation relations of charging and operating SAEV flows and passenger flows, ridesharing passenger number limit, and capacity limits of supporting facilities. To improve the solution efficiency of the model, linear approximation and linear equivalence techniques were employed to reconstruct the model into a mixed-integer linear programming model, and the Epsilon-constraint method was used to solve the Pareto-optimal solutions of the multi-objective model. The validity of the model was verified using the travel data of the road network in Chengdu, and a scenario comparison analysis was conducted for different numbers of ridesharing passengers and ratios of fleet charging demand. Research results show that when the number of ridesharing passengers increases from one to four, the total vehicle travel distance decreases by 77.87%; the fleet size decreases by 88.56%, and the construction cost of supporting facilities decreases by 96.80%, but the total passenger travel time increases by 125.46%, which indicates that operators should select an appropriate ridesharing strategy to ensure passenger travel efficiency. When the proportion of SAEV charging demand decreases from 30% to 5%, the total vehicle travel distance decreases by 10.77%, the operator fleet size decreases by 3.69%, and the construction cost of supporting facilities and total passenger travel time remain unchanged, which indicates that improving the SAEV endurance performance has great potential in improving the transport efficiency of SAEV fleet and reducing the operation cost of SAEV fleet.
2026, 26(6): 198-208. doi: 10.19818/j.cnki.1671-1637.2026.118
FullText HTML (26) PDF (4003KB) (8)
Dynamic path planning method considering load balancing of road network for AGV sorting system
LIU Zhi-shuo, XU Jun-zhe, LI Yan-hua, LI Xin
Abstract: More> To solve the path planning and road network congestion problems of automated guided vehicles (AGVs) in large-scale automatic sorting systems, by considering the load balancing and utilization rate of the road network, a path planning framework based on a sliding time window that integrated task-level global planning and action-level local adjustment was proposed. A multi-endpoint A* (A*-Ⅰ) algorithm considering the number of turns was designed to plan the global paths of AGVs entering the system. Based on the sliding time window framework, the position information of running AGVs in the system was updated to help them select actions to avoid path conflicts. By calculating the average passing speed of each road section in the road network every certain period, the road resistance factor matrix of each region in the road network was updated, and an A*-Ⅱ algorithm considering the road resistance factor was designed based on the A*-Ⅰ algorithm to adjust the local paths of AGVs. By combining the A*-Ⅰ and A*-Ⅱ algorithms, the conflict-free path planning of multiple AGVs in the automatic sorting system was realized, and the road network balancing and utilization rate of the sorting system were improved. Based on the cellular automata method, the operation and status update rules of AGVs in the system were determined, and a simulation framework for a large-scale AGV sorting system was constructed. Research results indicate that compared with the traditional global path planning method, the proposed path planning method reduces the load ratio of high-load nodes in the road network by 17% and the standard deviation of the overall road network load by 6.7%, effectively increasing the sorting quantity of goods per unit time. The road resistance factor weight and the road network status update cycle are the main factors affecting the system operation efficiency, and the system sorting efficiency is the best when road resistance factor weight is greater than 4, and road network status update cycle takes five time steps. The proposed method can effectively alleviate the road network congestion of large-scale AGV sorting systems, improve the system throughput and operation stability, and provide a feasible technical scheme for real-time path scheduling in intelligent warehousing and automatic sorting scenarios.
2026, 26(6): 209-220. doi: 10.19818/j.cnki.1671-1637.2026.120
FullText HTML (33) PDF (4929KB) (7)
Performance evaluation of post-earthquake material reception in airport hubs considering dynamic receiving efficiency
HOU Zong-hao, LI Gang
Abstract: More> To scientifically evaluate the emergency material reception capacity of airport hubs after earthquakes, the material receiving efficiency index (MREI) was proposed in response to the shortcomings of traditional static evaluation methods in terms of time-varying characteristics and system dynamic response. Combined with engineering system resilience theory and dynamic efficiency, a full-cycle evaluation framework for post-earthquake airport emergency material reception capacity was established. The dynamic evaluation of post-earthquake airport material reception capacity was realized by quantifying the time-varying deviation between actual and ideal cumulative material received based on MREI. The impact of equipment damage, dynamic scheduling, and resource interaction on receiving efficiency was reflected in real time. A multi-agent model was constructed to enable the quantitative calculation of MREI by simulating the impact of equipment damage, dynamic scheduling, and resource interaction on material reception. The model integrates a Bayesian network to quantify the functional status of subsystems, and adopts the Monte Carlo method to simulate seismic uncertainty. The case analysis shows that based on MREI, the impact law of peak ground acceleration, aircraft type ratio, and arrival interval on reception capacity can be effectively revealed. Post-earthquake airport material reception capacity decreases gradually with higher strength, and 4E airports are generally superior to 4D ones. When the proportion of large aircraft exceeds 0.7 (for 4E airports) and 0.4 (for 4D airports), the reception capacity attenuates significantly. For 4E airports with an ideal material reception demand of 2 000 t, there exists an optimal arrival interval range (16 - 19 min). This range allows MREI to be maintained above 0.8, and shortens the total material reception time by 23% compared with the minimum interval (3 min). MREI overcomes the limitations of traditional static indexes such as throughput and the number of flight diversions, and provides an integrated quantitative tool of "damage assessment-function calculation-dynamic simulation-strategy optimization" for post-earthquake emergency decision-making. Its dynamic evaluation capability also offers valuable methodological support for research on transportation hub resilience.
2026, 26(6): 221-238. doi: 10.19818/j.cnki.1671-1637.2026.028
FullText HTML (22) PDF (8244KB) (8)
Location-routing optimization for joint land-sea emergency delivery to large islands
LI Jia-cheng, WU Di, WANG Feng, LIU Bao-li, ZHENG Jian-feng
Abstract: More> Considering some real-world factors, such as material priority, fleet heterogeneity, and the personnel/material capacities of assembly ports and vessels, a location-routing optimization model was constructed. This model employed assembly port location, vessel voyage scheduling, and route configuration as variables, aiming to minimize delivery time. An integrated optimization algorithm was proposed. It leveraged an adaptive large neighborhood search algorithm as the framework for the outer-layer optimization loop. An improved simulated annealing algorithm was embedded as the inner-layer optimization module. The integration of global and local search was achieved through an interaction mechanism between the inner and outer layers. The effectiveness of the proposed model and algorithm was validated through an empirical case study of emergency delivery to China's South China Sea region. Research results show that the optimized delivery time is reduced from 78.52 h to 48.03 h, showing an efficiency improvement of 6.54%-48.51% and a stability enhancement of 10.77%-72.92% in comparison with existing algorithms. Further sensitivity analysis reveals that vessel capacity, speed, the number of assembly ports, and the average speed of ground transportation are all negatively correlated with delivery time of the system, while delivery personnel and material quantities, the number of landing ports are all positively correlated with delivery time of the system, exhibiting diminishing marginal effects. Deploying small or slow vessels can significantly undermine system performance. The proposed algorithm has the capability of effectively balancing material priority conflicts and heterogeneous transportation resource constraints. It provides timely and robust decision support for joint land-sea emergency delivery, and offers theoretical implications for extending the research on location-routing problems in special scenarios.
2026, 26(6): 239-256. doi: 10.19818/j.cnki.1671-1637.2026.119
FullText HTML (33) PDF (8932KB) (8)

Website copyright Journal of Traffic and Transportation Engineering Editorial Department陕ICP备05001904号-1

Address ：Editorial Department of Journal of Traffic and Transportation Engineering, Chang'an University, Middle Section of South 2nd Ring Road, Xi'an, China（710064） Tel：029-82334388 Email：jygc@chd.edu.cn

All visit:Today's visit:

Supported by: Beijing Renhe Information Technology Co. Ltd