交通运输工程学报

Cover and Contents of Vol.26, No.2, 2026

Cover Contents

2026, 26(2): .

FullText HTML (100) PDF (1601KB) (45)

Review of multimodal transport logistics engineering based on bibliometrics

YANG Zhong-zhen, ZHANG Er-zhuo, DU Yi-ying, XU Zhi-gang

Abstract: More> To systematically deconstruct the knowledge spectrum and research progress of multimodal transport logistics engineering, a structured analysis framework was constructed for the existing literature in the field of transport logistics based on the bibliometric method. By searching transport logistics-related literature from 1995 to 2024 throughout the CNKI database and Web of Science core database, 20 125 valid papers involving 5 948 authors and 5 969 keywords were selected. CiteSpace was applied to analyze the evolution trajectory and cutting-edge hotspots of the knowledge system of multimodal transport logistics. The results indicate that in terms of research perspectives, the Chinese publications mainly focus on urban logistics and reverse logistics, with attention paid to the model application in specific scenarios, while the English literature centers on maritime logistics and supply chain logistics, highlighting theoretical innovations. In terms of cooperation pattern, the Chinese publications form a study system dominated by the transportation-centered universities with a relatively loose inter-institutional cooperation network, while the English literature, with Asian maritime universities as the core, creates a cross-institutional, cross-regional, and close cooperation network. In terms of hotspot distribution, urban logistics studies focus on vehicle routing, maritime logistics on container liner operation, reverse logistics on closed-loop supply chain and green logistics, and site logistics on inventory control and facility layout optimization. In terms of research trends, the research hotspots in the field of transport logistics are evolving from the traditional technical optimization orientation to the emerging research direction centered on green development and supply chain resilience. Intelligence, greening, and resilience will become the main trend for multimodal transport logistics research in the future, thus promoting the logistics industry to a more efficient, environment friendly and flexible direction.

2026, 26(2): 1-23. doi: 10.19818/j.cnki.1671-1637.2026.071

FullText HTML (126) PDF (17190KB) (357)

Review of data-driven short-term prediction methods for continuous traffic flow in road networks

LIU Wei, ZHONG Can, CAO Wen-ming

Abstract: More> In the context of the rapid development of artificial intelligence computing and real-time traffic data acquisition technology, the deep learning prediction models, data processing technology, and prediction performance for short-term traffic flow were reviewed and summarized to grasp the latest developments of data-driven short-term traffic flow prediction technology for road networks. The evolution of classical statistical models, machine learning models, and deep learning models for traffic flow prediction was reviewed, and the advantages and limitations of various models were emphatically analyzed. The research progress of short-term traffic flow prediction methods from 2024 to the present was summarized. Short-term traffic flow prediction models such as recurrent neural networks, graph convolutional networks, multi-head attention mechanisms and Transformer architectures, neural ordinary differential equations, hypergraph theory, and lightweight architectures were compared and investigated in detail, as well as data processing technologies for short-term traffic flow prediction including federated learning, transfer learning, generative adversarial networks, and multi-source data fusion. Based on the comparison of three core indicators including root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE), the performance of mainstream models on the standardized dataset PEMS was summarized, and the generalization capability and stability of representative models were evaluated. The research results show that deep learning methods have significant advantages over traditional models in terms of accuracy, generalization capability, and stability for short-term traffic flow prediction. Short-term traffic flow prediction models with characteristics such as dynamic spatio-temporal relationship modeling, multi-scale periodic data structures, computational efficiency improvement methods, and enhanced robustness mechanisms demonstrate superior performance. Data processing technologies can effectively mitigate practical problems such as data privacy, cross-regional differences, data scarcity, and abnormal missing values, enhancing the engineering application performance and scalability of short-term traffic flow prediction models. Future research can be deepened in aspects such as spatio-temporal feature mining, data fusion, model light weighting, knowledge transfer, and model engineering applications.

2026, 26(2): 24-43. doi: 10.19818/j.cnki.1671-1637.2026.141

FullText HTML (240) PDF (3983KB) (246)

Method for identifying low-accessibility areas and their contributing factors in public transit accessibility to integrated passenger transportation hubs

CHEN En-hui, JI Ke-yu, CHENG Long, ZHANG Wen-bo, TENG Jing

Abstract: More> Regarding the challenge of identifying low-accessibility areas and their contributing factors in public transit accessibility to large-scale integrated passenger transportation hubs, a diagnostic framework was proposed for hub transit services based on accessibility. The spatial coverage and population served by multimodal networks were measured across different accessibility time thresholds. The Lorenz curve model was applied to assess the equity of spatial and population coverage across different hubs. A method was established to identify low-accessibility areas in hub public transit services, examining their spatial extent and distribution patterns. A gradient boosting decision tree model was introduced to analyze how various travel chain components affect hub accessibility from a behavioral perspective. An affinity propagation clustering algorithm targeting low-accessibility areas was designed to categorize distinct types of poorly connected zones. A case study was conducted using multimodal transportation networks at Hongqiao and Pudong hubs in Shanghai. Research results show that although Hongqiao hub generally has better public transit accessibility than Pudong hub, it performs worse in terms of equity, with more pronounced spatial disparities. Low-accessibility areas to Hongqiao exhibit a multi-core, dispersed pattern, while those to Pudong show a linear, clustered distribution. Regarding the contributing factors to low accessibility, walking distance has the greatest relative impact on Hongqiao hub (31%), followed by the road network detour index (29%) and the number of surface bus stops (21%). For Pudong hub, walking distance's influence increases to 37%, followed by the number of bus stops (26%) and the public transit network detour index (18%). Based on these key factors, the low-accessibility areas to both hubs were categorized into three types: first-and-last-mile walking-constrained, bus-dependent, and long-distance detouring rail transit.

2026, 26(2): 44-60. doi: 10.19818/j.cnki.1671-1637.2026.142

FullText HTML (137) PDF (9671KB) (245)

Multi-operator collaborative scheduling model for electric apron shuttle buses at hub airports

BAO Dan-wen, QU Yang, ZHOU Jia-yi

Abstract: More> In response to high allocation cost, low scheduling efficiency, and long delays caused by the independent operation of electric ferry buses at hub airports, a scheduling method under a multi-operator collaborative scheduling mode was proposed for electric ferry buses at airports. An operation mechanism was designed for operators to dispatch buses collaboratively, allowing buses to serve flight requests from different operators. A mixed-integer programming model was constructed for both independent and collaborative modes, aiming to minimize bus operation cost and the total delay in flight support. Two parameters, namely, shared radius and number of shared vehicles, were introduced to quantify the degree of collaboration. An improved genetic algorithm (IGA)-based solution framework was designed. Through the collaborative scheduling strategy, the overall airport operating cost was reduced and flight delays were mitigated. Empirical validation was carried out with Guangzhou Baiyun International Airport as an example. According to the results, compared with the traditional independent scheduling mode, the proposed multi-operator collaborative model significantly improves bus utilization. Under complex operational scenarios, delays can be eliminated by 69.4%–100.0% and operating costs decease by 21.6%–75.7%. The improved genetic algorithm shortens computation time by 85.3% to 99.5% compared to traditional solvers. A high-quality solution can be acquired within a shorter time, while the gap with the objective function can be maintained within 1%. For busy hub airports, the proposed method provides an effective balance between reducing bus operating cost and minimizing flight delays.

2026, 26(2): 61-78. doi: 10.19818/j.cnki.1671-1637.2026.143

FullText HTML (102) PDF (8336KB) (217)

Construction of evaluation index system for collaborative layout of port clusters and practical exploration of Shenzhen Port

HUANG Qi-xiang, JIANG Zhi-bin

Abstract: More> Different from the traditional collaborative evaluation model of port clusters with cities as the basic unit, a technical method was proposed for collaborative evaluation of port clusters with port areas instead of cities as the basic unit based on the actual organizational characteristics of port shipping supply chains. Specifically, the evaluation was conducted from two dimensions: collaboration among different port areas in cities and collaboration among different ports in regions. To effectively evaluate the issues in the collaborative layout of coastal cities within port clusters and accurately determine the level of internal and external collaborative development of ports, the "combined weighting-nonlinear gain" method was employed to construct a collaborative evaluation model based on the system collaboration theory. Taking Shenzhen Port as an example, the analytic hierarchy process (AHP) was used to invite expert scoring to construct a judgment matrix and determine the subjective weights. Based on the entropy weight method, the information entropy was calculated with the panel data of Shenzhen Port from 2018 to 2024 to determine the objective weights. The subjective and objective weights were integrated using the weight coefficient. A collaborative gain term was innovatively introduced. The main core indicators from the evaluation index system of city-level and regional-level port clusters were selected to construct a nonlinear evaluation model to quantify the internal and external collaborative effects of port cities. The empirical results show that in 2024, Shenzhen Port has achieved a collaboration score of 0.831 8 within the city-level port cluster, indicating an overall excellent collaborative state. However, there are significant differences among the five indicators. The information collaboration (0.90) is the best, while ecological and environmental collaboration (0.50) is the worst, followed by operational management collaboration (0.75), institutional collaboration (0.73) and infrastructure collaboration (0.71). Therefore, the port areas within Shenzhen Port should focus on strengthening energy sharing and infrastructure construction. In the Guangdong-Hong Kong-Macao Greater Bay Area port cluster, Shenzhen Port gains a collaboration score of 0.682 8, indicating a primary collaborative state. There is still considerable room for improvement in Shenzhen Port's collaborative layout within the regional port cluster. Among the six indicators, information collaboration (0.76) is relatively the best, operational management collaboration (0.66) reaches a primary level, while infrastructure collaboration (0.57), economic benefit collaboration (0.48), institutional collaboration (0.42) and ecological and environmental collaboration (0.35) show serious imbalances. Shenzhen Port should implement a system and ecology-driven strategy. Internally, based on an improved collaborative layout of infrastructure, the establishment of a port alliance revenue distribution mechanism should be prioritized to accelerate the construction of an LNG bunkering network and improve the core weaknesses within a certain period, effectively enhancing energy utilization efficiency. Externally, the collaborative safeguard policies for the Guangdong-Hong Kong-Macao Greater Bay Area port cluster should be further studied.

2026, 26(2): 79-93. doi: 10.19818/j.cnki.1671-1637.2026.144

FullText HTML (99) PDF (15561KB) (64)

Short-term passenger flow prediction framework for comprehensive transportation hub based on behavioral representation spatial correlation network

DAI Zhi-cheng, LI De-wei, LI Hua, XU En-hua, ZHANG Ruo-nan

Abstract: More> To improve the accuracy of short-term passenger flow prediction for heterogeneous functional areas within the transportation hub, the prediction accuracy issue caused by the limitation of existing prediction methods in accurately capturing complex spatiotemporal interactions among non-adjacent areas was investigated. The comprehensive application of virtual reality (VR) technology was studied and the multi-scale temporal feature encoding module was improved to establish a behavioral representation spatial correlation network (BRSCN) framework for short-term passenger flow prediction. A comprehensive transportation hub scenario was constructed using VR technology. The passenger behavior experiments were conducted to collect trajectory data of passengers in the virtual environment. The trajectory data were processed through regional identification, dwell-time detection, and activity-chain reconstruction to generate a complete set of passenger spatial activity chains. Subsequently, sliding-window sampling and graph embedding techniques were applied to derive feature vectors for each functional area. A spatial correlation graph reflecting non-adjacent inter-area relationships was constructed by jointly considering cosine similarity and passenger flow transition frequencies. During the prediction stage, a graph attention network (GAT) was employed to aggregate spatial features from both adjacent and non-adjacent areas. A multi-scale temporal feature encoding module was then designed by combining a bidirectional extended long short-term memory network (Bi-sLSTM) with a dynamic-window sparse-attention Transformer, enabling adaptive capturing of complex nonlinear and multi-scale spatiotemporal passenger flow fluctuation features within the hub. According to experimental results, compared with state-of-the-art models such as AGCRN and STTN, the real passenger flow data selected from the elevated level of the Shanghai Hongqiao comprehensive transportation hub reduces root mean square error (RMSE), mean absolute error, and mean absolute percentage error by 30.2%, 21.1%, and 28.3%, respectively. The incorporation of the spatial correlation graph further decreases RMSE by 16.7%, significantly enhancing the model's ability to predict passenger flow interactions between non-adjacent areas. Moreover, without lowering prediction accuracy, the dynamic-window sparse attention mechanism reduces model complexity by 2.9%. The proposed BRSCN framework effectively captures complex spatiotemporal relationships among non-adjacent functional areas within comprehensive transportation hubs, substantially improving short-term passenger flow prediction accuracy and model generalization performance. A scientific decision-making basis can be provided for spatial resource allocation optimization and dynamic passenger flow management in comprehensive transportation hubs.

2026, 26(2): 94-109. doi: 10.19818/j.cnki.1671-1637.2026.026

FullText HTML (109) PDF (15410KB) (86)

Multi-layer heterogeneous take-off and landing site network collaborative planning for urban low-altitude logistics

FENG Di-kun, ZHANG Hong-hai, HUA Ming-zhuang, DI Juan, SHAN Yi-xuan

Abstract: More> To achieve effective planning and safe operation of take-off and landing site networks for low-altitude logistics, the collaborative planning of multi-layer take-off and landing site networks for unmanned aerial vehicle (UAV) logistics in urban environments was investigated. By considering realistic factors such as UAV performance constraints and airspace restrictions and combining a three-layer distribution system of "vertihub-vertiport-terminal", a three-layer network collaborative planning model for take-off and landing sites was built to achieve minimum transportation cost, minimum construction quantity and maximum network fairness. A customized continuous optimization framework was designed to quickly realize the generation of initial solutions, solution of discrete variables and local optimization of continuous variables. The effectiveness of the collaborative planning model and the combined algorithm was validated based on logistics data from Shanghai. The results show that the proposed collaborative planning model can effectively realize the location layout of the take-off and landing site network. The designed HLO (Human Learning-based Optimization) algorithm performs better than optimization algorithms including GA (Genetic Algorithm) and NSGA-Ⅱ (Non-dominated Sorting Genetic Algorithm Ⅱ). Significance tests indicate that the SA-based local optimization algorithm can stably improve transportation cost by approximately 2.15%, effectively enhancing the optimality of planning results. Sensitivity analysis reveals that the weight design and step factor significantly affect the solution effect of the model. The weight of the total number of vertihubs and vertiports should not be less than 0.5, and the optimal step factor is between 0.05 and 0.06. The weight configuration for the number of vertihubs and vertiports notably affects the number of vertiports. The number of vertiports continuously increases with its weight, while the optimal number of vertihubs remains unchanged. The research can provide decision-making reference for network planning and collaborative operation of take-off and landing sites in urban low-altitude logistics.

2026, 26(2): 110-124. doi: 10.19818/j.cnki.1671-1637.2026.086

FullText HTML (63) PDF (7772KB) (41)

Optimization method of customized bus feeder routes at comprehensive transport hubs under ride-hailing collaboration

CHENG Long, WANG Yu-xuan, XIAO Guang-han, ZHANG Ming-ye, YANG Min

Abstract: More> A ride-hailing coordinated customized bus feeder service mode for comprehensive passenger transport hubs was proposed. The minimization problem of total system cost under the coordination of customized buses, ride-hailing services, and passengers was studied. Influencing factors including the distribution of passenger feeder travel demand, spatial connectivity of bus routes, time scheduling, passenger flow assignment, and walking accessibility were considered. An optimization model for customized bus feeder routes under ride-hailing collaboration was established. An embedded optimization algorithm based on large neighborhood search was proposed. The mixed-integer linear programming model was embedded into the large neighborhood search framework. Global optimization of customized bus feeder routes was achieved through a three-step strategy consisting of perturbation, repair, and exact optimization. An empirical study was conducted with Nanjing Lukou International Airport as a case. The costs and operational performance under the coordinated service mode and single service modes were systematically evaluated. The results show that the proposed model can effectively satisfy passengers' main feeder travel demand and significantly reduce the total cost of the feeder system. Compared with the single ride-hailing mode and the single customized bus mode, the optimized total cost of the feeder system decreases by 40.7% and 18.8%, respectively. Sensitivity analysis shows that the improvement of customized bus operating speed has a more significant impact on the total system cost than ride-hailing speed. The total system cost shows slightly higher sensitivity to changes in ride-hailing fare than to customized bus fares. Passenger travel cost is the main factor influencing changes in total system cost. When the acceptable walking distance for passengers is 700 m, the total system cost reaches the minimum value. Algorithm comparison results indicate that the embedded optimization algorithm based on large neighborhood search is superior to genetic algorithms, simulated annealing algorithms, ant colony algorithms, and traditional large neighborhood search algorithms in both the number of iterations and solution quality. The embedded optimization algorithm further reduces the total system cost by 9.6% to 12.6%.

2026, 26(2): 125-139. doi: 10.19818/j.cnki.1671-1637.2026.146

FullText HTML (86) PDF (7187KB) (30)

Optimization model for trunk-regional-general hierarchical air transportation network with stratified demand

JIANG Yu, LIU Hong-yang, LI Han-lu, YUAN Ye, XUE Qing-wen

Abstract: More> The trunk-regional-general hierarchical air transportation network was taken as the research object, and modeling consistent with its operational characteristics was conducted to construct and solve a hierarchical hub location model with stratified demand. The demand layer was employed to distinguish various demands. To minimize the total cost, including transportation costs, hub construction fixed costs, and fixed costs for establishing flight connections, an r-allocation hierarchical hub location model permitting non-hub direct connections and regular hub direct connections was built. According to the topological characteristics of air transportation networks, a VNS-GA hybrid heuristic algorithm based on an alternating mechanism was designed by combining the advantages of the VNS algorithm and the genetic algorithm (GA). Hub selection and demand node allocation were optimized by VNS, while direct connections were optimized by GA. The modeling and solution were carried out for the two classical datasets, namely, Civil Aeronautics Board (CAB) and Australia Post (AP), as well as the Yangtze River Delta regional airport data in China. The existing model and the stratified demand model were compared to verify the effectiveness of the algorithm and analyze the parameter sensitivity. According to the research results, in the small-scale case of 15 nodes, the stratified demand model reduces the total cost by 9.23%. In the small-scale case of 25 nodes, the alternating VNS-GA algorithm has a gap with the optimum solution of no more than 2.56% under various parameter configurations, with the average solution time only 10.78% of that of the commercial solver. In the medium and large-scale case of 100 nodes, sensitivity analysis shows that the setting of stratified weight coefficients has a great impact on the optimization results. The r-allocation strategy can reduce the total cost but with obvious diminishing marginal benefits. In the semi-empirical experiment of the Yangtze River Delta region, the model can reduce the cost by 2.75% while adding 50 direct connections. This verifies the feasibility and effectiveness of the model in optimizing trunk-regional-general hierarchical air transportation networks.

2026, 26(2): 140-154. doi: 10.19818/j.cnki.1671-1637.2026.147

FullText HTML (68) PDF (8425KB) (22)

Resilience evaluation method of external transportation network of integrated transit hub cluster under time-varying demand

YANG Yang, CHEN Jun-ting, YAO En-jian, LIU Dong-mei

Abstract: More> To scientifically evaluate the resilience of the external transportation network of the integrated hub cluster, a travel resilience measurement indicator was proposed, covering three dimensions: robustness, redundancy and resilience. The multimodal traffic network passenger flow assignment method was employed to characterize the influence results of different node attack and recovery strategies. The travel resilience measurement method was constructed. Taking the integrated transit hub cluster in the Beijing-Tianjin-Hebei urban agglomeration as an example, an example verification was carried out. Research results show that, the external comprehensive transportation network has been basically formed between the Beijing-Tianjin-Hebei urban agglomeration and the core cities of typical urban agglomerations such as Shanghai, Guangzhou, and Chengdu. The overall anti-interference and recovery capabilities are strong. The unit travel resilience is generally higher than 0.95 under different disturbance scenarios. Taking a single-node recovery time of 1 h in October 7th (National Day) as an example, the network recovery effect of the node strength recovery strategy is the best, with the network resilience of 0.990 2 and only 140 2 people affected, which is significantly superior to other strategies. In the morning peak period, due to the large network load and low redundancy of shifts, the resilience decreases to 0.976 1, lower than that in the afternoon and evening peaks. In most cases, the network resilience reduces with the longer single-node recovery time, while the number of affected people also increases. For different dates, resilience is affected by both the dispersion of passenger flow distribution and network redundancy. According to the TOPSIS method, the highest resilience is 0.983 2 on National Day, while the lowest resilience is 0.974 6 at weekends. The proposed resilience evaluation method can provide a scientific basis for analyzing the potential traffic capacity of multimodal traffic networks such as inter-hub railways, civil aviation and urban transport, and rationally allocating transportation emergency response resources.

2026, 26(2): 155-169. doi: 10.19818/j.cnki.1671-1637.2026.148

FullText HTML (82) PDF (6219KB) (19)

Simplified analytical solution for annual periodic temperature field along radial depth of lining and surrounding rock in deep-buried tunnels in cold regions

ZHAO Peng-yu, CHEN Jian-xun, LUO Yan-bin, YANG Dian-guang, FAN Wen-jie, LIU Ji-xiang

Abstract: More> An analytical solution with higher applicability was expected to be obtained for the temperature field along the radial depth of the lining structure and the surrounding rock in deep-buried tunnels in cold regions, so as to guide the design of tunnel frost damage prevention. Based on the superposition principle, the tunnel was simplified as a multi-layer cylinder calculation model. After clarifying the determination method for the temperature influence boundary of the surrounding rock, the annual average air temperature and initial surrounding rock temperature were taken as constant temperature boundary conditions to propose an analytical solution for the annual average temperature along the different radial depths of the tunnel lining structure and the surrounding rock. In addition, the tunnel was simplified into a multi-layer slab calculation model. The temperature harmonic wave propagation principle was applied. The effects of thermal property differences between the lining concrete and the surrounding rock on the propagation of the temperature harmonic wave were considered. Analytical solutions of the annual temperature amplitude along the different tunnel radial depths of the lining structure and the surrounding rock were proposed. According to the analysis results, the functional form of the analytical solutions for the annual average temperature and annual temperature amplitude along the different radial depths of the tunnel lining structure and the surrounding rock is consistent with that of the mathematical characterization of the temperature field variation patterns obtained through field tests and numerical simulations. The annual average temperature and annual temperature amplitude vary logarithmically and exponentially along the radial depth, respectively, with the approximately linear variation of the lining structure. The gap between the theoretical analysis and numerical simulation results of the annual average temperature is small, which is lower than 0.1 ℃. The theoretical analysis results for the annual temperature amplitude are larger than those obtained from numerical simulation calculations, which are relatively conservative. The maximum gap between the two can reach up to 1.3 ℃, occurring at a radial distance of approximately 3 m from the lining surface. The gap between theoretical analysis and numerical simulation results for the annual temperature amplitude is most significantly influenced by the annual air temperature amplitude and the thermal conductivity of the surrounding rock. As the annual air temperature amplitude and the thermal diffusivity of the surrounding rock decrease, the gap gradually narrows. When the annual temperature amplitude is less than 15 ℃ and the thermal diffusivity of the surrounding rock is below 1.10×10^-6 m²·s^-1, the gap between the two remains below 1 ℃.

2026, 26(2): 170-185. doi: 10.19818/j.cnki.1671-1637.2026.149

FullText HTML (73) PDF (4402KB) (23)

Calculation of displacement response induced by tunnel excavation considering restraint effects of in-service piles and soil stiffness hardening

HUANG Ming, ZHANG Cheng-zhao, LAI Feng-wen, LU Yao, WEI Wei

Abstract: More> Neglecting the restraint effect of in-service piles and the stress-dependency of soil stiffness in pile-soil interaction may lead to a certain degree of discrepancy between the theoretically calculated values and the realistic values of displacement responses induced by tunnel excavation. Therefore, a two-stage analysis method (TSAM) was adopted to develop a pile-soil interaction model incorporating both the restraint effect of in-service piles and soil stiffness hardening. Moreover, an equivalent internal friction angle integrating soil cohesion was introduced to correct the Loganathan-Poulos (L & P) formula and calculate soil displacements induced by tunnel excavation adjacent to in-service piles. Using a Winkler model with nonlinear soil springs, the displacement of existing piles resulting from tunnel excavation was subsequently determined. Finally, the proposed solution was utilized to explore the rules of pile displacement responses induced by tunnel excavation, taking into account the influence of pile-tunnel spacing, ground loss ratio, tunnel depth, and pile diameter. The results demonstrate that the proposed solution exhibits excellent agreement with the existing theoretical solutions, centrifuge test data, and field monitoring results for surface settlement, offering a significant improvement over the traditional L & P formula. The change law of in-depth soil displacement solution shows strong consistency with that of existing theoretical solutions. Notably, the pile displacements predicted by considering the restraint effect of in-service piles and soil stiffness hardening are smaller than those estimated by existing theoretical solutions. Under different ground loss ratio scenarios, the calculated pile displacement exhibits consistent overall trends with finite element (FE) results. At the tunnel axis elevation, the calculated vertical pile displacements are about 2% lower than the FE results, with horizontal displacements being 10% smaller than the FE results.

2026, 26(2): 186-197. doi: 10.19818/j.cnki.1671-1637.2026.009

FullText HTML (54) PDF (2392KB) (19)

Field test on horizontal bearing characteristics of large-diameter PRC pipe piles installed with pre-drilled inserted method

REN Yu-bo, WANG Miao-miao, TIAN Ye-qing, XIE Hong-li, MA Kai, ZHOU Zhi-jun, XU Chao

Abstract: More> To study the horizontal bearing characteristics of large-diameter prestressed reinforced concrete (PRC) pipe piles installed with pre-drilled inserted method, the field static horizontal load tests were conducted on PRC pipe piles with an 800 mm diameter. The differences in displacement, internal forces, and proportional coefficient of foundation resistance coefficients of PRC pipe piles with pre-drilled inserted method and hammer-driven method were compared. The bearing mechanisms of PRC pipe piles with the two methods were also analyzed. The results show that compared with hammer-driven pipe pile, the horizontal critical load of pre-drilled inserted pipe pile increases by 36.1%. Their ultimate load is similar. Under the critical load, the rotation point of the pile shaft is 7.5 m for hammer-driven pipe pile and 5.5 m for pre-drilled inserted one. The depths of the maximum bending moment section and the maximum soil resistance section coincide, being 5.1 m for the hammer-driven pipe pile and 3.1 m for pre-drilled inserted one. When the horizontal load is less than the critical load, the pre-drilled inserted pipe pile exhibits smaller bending moments and greater lateral soil resistance than the hammer-driven one. The proportional coefficient of foundation resistance coefficients of PRC pipe pile with the two methods exceeds the recommended values in codes, with the pre-drilled inserted pipe pile showing more significant improvement. The m-method demonstrates good applicability in calculating horizontal load effects for pre-drilled inserted pipe pile, but is unsuitable for hammer-driven pipe pile. During the implantation of pre-drilled inserted pipe pile, the upward migration of mortar along the borehole wall fills and penetrates the soil. The bonding capacity between the mortar and soil is enhanced, the horizontal resistance of the foundation soil is increased, and a better displacement control capability is provided compared to hammer-driven pipe pile, indicating a promising application potential in structures with strict displacement requirements, such as bridges.

2026, 26(2): 198-209. doi: 10.19818/j.cnki.1671-1637.2026.010

FullText HTML (33) PDF (5733KB) (18)

Experiments on shear properties of vetiver root-red clay root-soil composite under different failure modes

FANG Wei, LIU Huo, LIU Tian, WANG Yi-wei

Abstract: More> To investigate the influence of vetiver roots on the shear properties of red clay, one-year-old vetiver root-red clay composites were selected as the research object. After obtaining the basic parameters of red clay and vetiver roots, the shear strength and shear characteristic parameters of the root-soil composite under different failure modes were obtained through direct shear tests, single root pullout tests, and interfacial shear resistance tests. Furthermore, the variation laws of shear strength under the influences of root segment inclination angle, soil dry density, and water content were analyzed. The results show that the direct shear strength of the root-soil composite under different normal loads is greater than that of plain soil. When the angle between the root growth direction and the shear direction is smaller, the increase in soil shear strength is larger. When the root axis is parallel to the shear direction, the root reinforcement effect is better than that under orthogonal conditions. The addition of roots shows a significant effect on the improvement of soil cohesion. The pullout force increases with increasing pullout displacement and reaches a peak at a displacement of 3-8 mm, after which two stages of rapid decrease and slow decrease are experienced until the root segment is completely pulled out. Both the pullout force and the interfacial shear strength exhibit positive correlations with soil dry density, but peak values appear when the soil water content is approximately 20%, that is, near the optimum water content. With increasing soil dry density, the bonding between the root surface and soil particles is enhanced, while the root-soil contact area and friction effect increase simultaneously. The interfacial shear strength, interfacial cohesion, and interfacial friction coefficient ratio of the root-soil interface all reach maximum values at a water content of approximately 20%. Under a given level of soil compaction, the root-soil friction effect can be improved and root pullout failure can be reduced by regulating the humidity of the root growth environment. The research results provide experimental evidence for the accurate evaluation of the shear strength of root-soil composites under different failure modes and have reference value for the reasonable selection of shear parameters in root-reinforced soil models.

2026, 26(2): 210-224. doi: 10.19818/j.cnki.1671-1637.2026.007

FullText HTML (69) PDF (6218KB) (13)

Efficient image segmentation method for interface heterogeneous materials of grouted asphalt concrete under small sample conditions

BAI Tao, ZHAN Ao-yang, GU Fan

Abstract: More> An intelligent segmentation framework coupling physical constraints and attention mechanisms was proposed. An adaptive erosion-dilation preprocessing algorithm was proposed, combining with multi-weight edge fusion operators to enhance gradient responses of interfaces. The atrous spatial pyramid pooling-squeeze and excitation (ASPP-SE) network architecture was designed to achieve multi-scale feature decoupling of pore distribution and aggregate texture through synergistic optimization of SE-Block and ASPP. A joint correction strategy integrating edge confidence and segmentation masks was established to optimize initial segmentation results through post-processing. The results demonstrate that on our self-constructed small-sample dataset of grouted asphalt concrete, the proposed method achieves an improvement of 8.4% in accuracy, 6.6% in precision, and 6.9% increase in recall compared to the SegFormer deep learning model, effectively enhancing boundary segmentation accuracy. This method effectively resolves segmentation failures in complex scenarios including blurred material phase interfaces and aggregate texture features, demonstrating superior reliability and generalizability in practical engineering applications. It also provides new insights for analyzing engineering material images with limited data and high heterogeneity.

2026, 26(2): 225-242. doi: 10.19818/j.cnki.1671-1637.2026.008

FullText HTML (58) PDF (19201KB) (18)

User Center

Journal Info

Index byMore>

Current Issue
2026, Volume 26, Issue 2

Search Type

Search Journal

LinksMore>