Abstract:【Objective】To establish the relationship between the deterministic safety design and reliability design of pile plate structure subgrade.【Method】This paper first derives the analytical solution for settlement calculation of pile plate structure subgrade and establishes the reliability analysis model for the subgrade. The influence of the variability coefficient of subgrade soil elastic modulus, upper load, and pile spacing on the failure probability of subgrade settlement is studied. Subsequently, a framework for settlement safety criteria of pile plate structure subgrade based on the generalized reliability ration of safety margin is proposed. The relationship between the relative safety margin of the generalized reliability index and the safety factor's relative safety margin is analyzed,and the effect of overlying load on the critical pile spacing is discussed.【Result】The results show that the settlement calculation method in this paper can better reflect the settlement deformation characteristics of the pile plate structure subgrade. The settlement of the pile plate structure subgrade is jointly determined by the upper load and the combined stiffness of the plate, and its failure probability curve with the load level changes from a steep increase to linear growth as the variability coefficient of the subgrade soil elastic modulus increases. The relative safety margin of the generalized reliability index and the safety factor's relative safety margin exhibit an approximately linear relationship. In addition, an increase in overlying load will lead to a logarithmic non-linear decrease in the critical pile spacing.【Conclusion】By allowing the failure probability to calibrate the allowable safety factor, the basic consistency of deterministic design and reliability design of the pile plate structure subgrade is ensured.

Abstract:【Objective】The overall elastic buckling performance of corroded cold-formed thin-walled steel columns was investigated in this paper.【Method】The finite element eigenvalue buckling analysis method was used to study the effects of corrosion location, corrosion area length and width on the elastic overall buckling of coldformed thin-walled steel columns. The calculation method for the elastic overall buckling stress of corroded coldformed thin-walled steel columns was proposed, and the theoretical calculation results were compared with the finite element eigenvalue buckling analysis values.【Result】The results show that as the degree of corrosion increased, the elastic bending buckling stress and the elastic bending torsional buckling stress gradually decreased. The effect of web corrosion on elastic bending buckling stress was greater than that of flange corrosion,while the effect of web corrosion and flange corrosion on elastic bending torsional buckling stress was similar. The elastic bending buckling stress gradually decreased with the increase of the corrosion area length, and the rate of decrease showed a decreasing trend. However, the elastic bending torsional buckling stress approximately linearly decreased with the increase of the corrosion area length.【Conclusion】With the increase of the corrosion area width, the elastic bending torsional buckling stress decreased faster than the elastic bending buckling stress.

Abstract:【Objective】The purpose of this paper is to address the issue of excessive axial force in the first support of a narrow and long deep excavation of metro stations during construction.【Method】Through the method of finite element analysis using the hardening soil model with small strain stiffness (HSS), the excavation method of the station pit , poor quality of steel support installation, and the untimely stressing of the steel support are simulated in a refined way to analyze the influence on the axial force of the first support.【Result】The results show that the use of basin excavation has a significant effect on the axial force of the first support, which increases by 45% compared with the case of layered excavation, and the corresponding position of the steel support should be installed in time; when the prestress of the installed steel support does not reach the design value, it will lead to an increase in the axial force of the top support at the end of the excavation, with an increase of 10% or less, and a decrease in the axial force of the bottom support.【Conclusion】Basin over-excavation should be strictly limited during excavation of narrow and long deep excavation of metro stations; The effect of temperature changes on support axial forces cannot be ignored; when the prestressing force of the installed lower steel support fails to reach the design value, it will lead to a large axial force of the top support at the bottom of excavation.

Abstract:【Objective】In order to explore the structural vibration and noise characteristics of high-speed railway CRTS Ⅲ type plates and double block tracks.【Method】Based on the method of finite element, multi-body dynamics and boundary element, the dynamic model and boundary element model of high-speed train-ballastless track-box girder coupling system are established to study the structural dynamic response and noise radiation characteristics of the train-track-box girder coupling system.【Result】The results show that the maximum wheel load reduction rate of the train is 0.053 6 and 0.165 7, and the deflection of the lower box girder is 1/9 673 and 1/8 457, respectively, when the train passes the CRTS Ⅲ plate and the double-block track structure section on the bridge. Compared with the double-block ballastless track, the displacement response of the track plate and the base of the CRTS Ⅲ plate is smaller, and the vertical vibration attenuation of the track-box girder system is faster, and the vibration damping performance is better. In addition, the total sound pressure level of the box girder noise of CRTS Ⅲ segment is lower than that of the two types of track structure. However, the difference between the peak noise of the far field point under the bridge is small, and the peak frequency is similar, both of which are about 31.5 Hz. The peak frequency under the bridge has large differences, the peak frequency of type Ⅲ board being 63 Hz, and the double block type being 20 Hz.【Conclusion】CRTS Ⅲ plate type ballastless track is more suitable for use in high-speed railway bridge sections.

Abstract:【Objective】The shear tests of four rectangular reinforced concrete (RC) beams were conducted to investigate the effects of the end-anchorage method and preloading on the shear behavior of the carbon fiber reinforced polymer (CFRP)-strengthened RC beams.【Method】The failure mode, load-deflection curve, characteristic load, strain developments of the stirrup and CFRP were analyzed. Finally, the shear capacity of the specimens was predicted using the existing calculation methods for the shear capacity of the CFRP-strengthened RC beam. 【Result】The test results show that the using U-shaped CFRP strengthening can significantly enhance the shear behavior of RC beams under the reliable end-anchorage. The shear strengthening effect is significantly affected by the end-anchorage methods. The mechanical anchorage is obviously superior to the CFRP sheet anchorage. Compared with the unstrengthened beam, the shear capacity of the strengthened beam with the mechanical anchorage increases by about 60%, whereas that with the CFRP sheet anchorage only increases by 9.8%. The preloading significantly decreases the cracking load of the strengthened beams, but does not affect the shear capacity.【Conclusion】The research results can provide reference for the engineering application of shear strengthening of RC beams with CFRP sheets.

Abstract:【Objective】In order to simplify the test process and improve the detection efficiency, a two-span continuous beam bridge in the actual project was taken as the research background, and the light load numerical analysis was carried out based on the moving vehicle, and the accuracy and feasibility of the evaluation method of the damaged bridge bearing capacity under vehicle-bridge coupling vibration were explored.【Method】The vehicle-bridge coupling vibration model was established, the deflection and strain dynamic time-history response of the control section of the bridge structure were extracted, the dynamic components were removed by empirical mode decomposition (EMD) and variational mode decomposition (VMD) and the quasi-static response curve was identified, and the influence of bridge deck roughness and vehicle speed on the accuracy of the static response recognition was analyzed. The loading efficiency of different loading positions was calculated by the different responses of the control section during vehicle driving, and the standard check coefficient of the quasi-static response of the control section of the damaged bridge under light load efficiency was calculated by using the “Changding model”, and the bearing capacity of the damaged continuous beam bridge was evaluated.【Result】Compared with EMD, the quasi-static response curve of VMD decomposition is in better agreement with the static response curve. With the increase of bridge deck roughness and the moving speed of the loading vehicle, the error of the recognized quasi-static response and the static response also gradually increases. The quasi-static response identified during the loading process of the moving vehicle is relatively accurate and can be calculated as the measured value of static load. The calculated standard check factors are all greater than 1, which accurately evaluates the damage state of the bridge.【Conclusion】It is highly applicable and feasible to evaluate the bearing capacity of damaged bridges based on vehicle-bridge coupling vibration.

Abstract:【Objectives】The existing prediction formula of vibration source intensity is only applicable to normal-speed subways, and cannot be applied to the vibration source intensity characteristics of fast underground lines with a maximum speed of 160 km/h.【Method】The vibration characteristics of multiple sections of a coastal city urban expressway were measured on site. The influence of vehicle speed, tunnel buried depth, curve and other factors on the vibration source strength was studied through data analysis.【Result】The vibration source intensity was between 71 dB and 80 dB at the speed of 90 km/h to 160 km/h in the straight (curve) line section. The vibration source intensity also increased with the increase of vehicle speed,and the vibration source intensity increased by 3.72 dB for every 20 km/h increase in speed. The vibration source intensity decreases with the increase of the buried depth of the tunnel. For every 10 m increase in the buried depth of the tunnel, the vibration source intensity decreases by 4.9 dB. In the same case, the vibration source intensity of the curve section (curve radius 1 850 m) is stronger than that of the straight section, and the value is about 7 dB. One year after the train was put into operation, due to the deterioration of the wheel-rail state, the vibration source intensity increased, and the difference between different trains was large. The monitoring data of seven months of a certain section were selected and analyzed. The vibration source intensity was relatively stable after the operation period was stable.【Conclusion】The vibration source intensity of the fast underground line is affected by many factors: it is linearly positively correlated with the train speed, the tunnel depth is linearly negatively correlated, the vibration source at the curve is stronger than that at the straight line, and the wheel-rail deterioration will also increase the vibration source intensity.

Abstract:【Objective】Aiming at the potential risk caused by overhead contact wire defects during railway operation, an improved YOLOv7x method for overhead contact wire defect identification is proposed.【Method】Firstly, Swin Transformer network is introduced at the end of the backbone feature extraction layer to replace the original extended and efficient layer aggregation network module, so as to improve the ability of the network to grasp global information. Then the SIoU(SCYLLA-IoU) loss function is used to replace the original network loss function, and the direction penalty mechanism is added to the convergence process of the prediction frame. Finally, CA is integrated with the extended and efficient layer aggregation network module to enhance the global receptive field of the neck network module.【Result】Experimental simulation results show that the accuracy of the model trained with the improved algorithm reaches 95.9%, which is 4.7% higher than that of the original YOLOv7x algorithm, and the detection speed reaches 52 frames per second.【Conclusion】The improved algorithm solves the problem of low detection efficiency in hanging strings defect identification, which may improve the efficiency of detection of hanging strings defect in practice.

Abstract:【Objective】To analyze the influencing factors and patterns of electric vehicle group charging and discharging decisions in the process of vehicle network interaction.【Method】In order to analyze the influencing factors and laws of electric vehicle group charging and discharging decisions in the process of vehicle network interaction, a decision evolution game model among different individuals within the electric vehicle group was established and the replication dynamic equation was improved by introducing user preference influence factors, so as to analyze the process of the evolution of electric vehicle group strategies the impact of various factors on the interaction strategies of electric vehicle groups.【Result】Simulation shows that when considering the influence of user preferences, the proportion of electric vehicle users who choose charging strategies increases from 65% to 75%. When the sensitivity of user economic benefits increases from 0 to 2, the proportion of charging and discharging users decreases by 50% and 10% respectively, while the proportion of discharging users increases by 60%.【Conclusion】The sensitivity of user preferences and economic benefits will increase the proportion of electric vehicle users participating in vehicle network interaction, while external environmental disturbances will not affect the final selection results of electric vehicle users.

Abstract:【Objective】In order to enhance the 3D coverage capability of wireless heterogeneous sensor networks (HWSN), a 3D deployment method based on improved honey badger optimization algorithm (IHBA) is proposed.【Method】Firstly, combined with the adaptive fruit fly optimization algorithm, the random search ability of the algorithm is enhanced, which is convenient for the algorithm to obtain the global optimal solution. Then, the strategy of replacing the worst individual is introduced to avoid the low adapting individual to occupy the population position and improve algorithm convergence speed. In the meantime, new individuals are introduced to improve the population diversity and avoid the individual prematurity of the algorithm.【Result】This algorithm is applied to the coverage optimization of HWSN, and the network coverage is improved by 18.1% compared with the standard Honey Badger algorithm.【Conclusion】The simulation results show that the algorithm converges faster, can effectively improve the network coverage capability of wireless heterogeneous sensors, and the node distribution of the whole network is more uniform.

Abstract:【Objective】As an important connecting part of new energy vehicles, special vehicles, locomotives and other means of transportation, rubber bushing plays a crucial role in the stability of the whole vehicle handling. In order to avoid the tedious work of repeated debugging in engineering practice,【Method】an optimal design method of rubber bushing all-directional stiffness (radial, axial, yaw and torsional stiffness) taking into account the material hardness and diameter reduction is proposed. Firstly, the Ogden model is used as the constitutive model of rubber bushing. Second, by combining experimental design with finite element analysis, the isotropic stiffness of rubber bushing with different inner core outer diameter, rubber height, diameter reduction and material hardness is calculated. The second-order response surface model of isotropic stiffness is established and the accuracy of the model is verified by Latin hypercube sampling and variance analysis. Finally, a multi-objective optimization method of isotropic stiffness of rubber bushing is proposed using genetic algorithm to testify the optimization result.【Result】The result shows that the relative errors between the measured radial, axial, yaw and torsional stiffness and objective stiffness are 7.72%, 9.06%, -6.33% and 9.16% respectively, which are all within ±10% and meet the requirements of engineering application.【Conclusion】The validity of the established second-order response surface model of isotropic stiffness and the feasibility of the proposed optimization design method are verified to provide guidance for the product design of rubber bushing, which can greatly shorten the research and development cycle of rubber bushing.

Abstract:【Objective】To solve the problems of selecting grinding modes based on experience and unpredictable grinding results in the current grinding operation by rail grinding trains.【Method】This paper takes the GMC-96X rail grinding train as the research object, to research the effects of different grinding modes by means of numerical simulation through MATLAB. By combining the distribution of grinding wheels in the grinding train, the basic model library of rail grinding is established, and a design method of rail grinding mode based on the grinding amount curve is proposed, which realizes the automatic generation of grinding mode and the precise repair of the rail head profile.【Result】The field test results show that the grinding mode designed by this method includes all angles of each grinding wheel of the grinding train, that can be directly input to the control system, which is more suitable for practical applications.【Conclusion】By comparing the measured rail profile before grinding, after theoretical grinding, and after actual grinding, the reliability of the method is verified.

Abstract:【Objective】In order to solve the problem of low magnetic field utilization of traditional magnetorheological brake (MRB), a drum MRB with internal and external fluid flow channel was designed.【Method】By adding non-magnetic rings and non-magnetic disk to the magnetic material rotary sleeve and the stator magnetic cylinder, the magnetic flux was guided meander through six effective damping gaps in the internal and external axial flow channels. Therefore, the torque performance was improved while keeping the outer dimension of the brake remains unchanged. The structure and working principle of the internal and external fluid flow MR brake were described, and the mathematical model of braking torque were deduced and established. Based on the analysis of electromagnetic field and torque, the accuracy of the model was predicted through theoretical calculation and DOE experiment orthogonal method, and the multi-objective optimization of the MR brake with internal and external fluid flow channels was carried out by using the NSGA-Ⅱ algorithm.【Result】The results show that,with an applied current of 2.0 A, the braking torque of the initial and optimal MRB dampers are 36.38 N·m and 47.35 N·m, respectively, which improves by 30.15%. Compared with the initial damper, the braking torque increased by 21.31 from 18.28, and the dynamic adjustable range improves by 16.58%.【Conclusion】The optimal MRB meets the braking performance requirements of unmanned delivery vehicles.

Abstract:【Objective】Hypergraph neural networks (HGNN) have the ability to learn inter-class uniqueness and intra-class commonality, which can significantly improve learning performance. However, traditional HGNN methods are in lack of the strong relational induction which determines the way how low-dimensional data nodes interact with each other. In order to solve this problem, a fuzzy HGNN(FHGNN) classification algorithm based on fuzzy theory is proposed, and hypergraph structure is constructed according to the characteristic information of data nodes.【Method】FHGNN first adopts an edge-focused GNN to make edge prediction through iterative updates of edge labels. The fuzzy membership function is designed according to the output of edge prediction to achieve a more accurate representation of the connection relationship between nodes. Finally, the hypergraph is constructed by the relation representation. Then the nodes are classified again and the result is obtained. The edge label loss function and node label loss function are used in FHGNN and their parameters are trained and learned respectively.【Result】Experimental results prove the proposed FHGNN method is more suitable for small-scale data with low node information dimension, and performs well in node classification tasks.【Conclusion】For classification tasks of different data sets, FHGNN can learn the relevant feature information of nodes more effectively and improve the learning effect.