Abstract:Taking an existing large-span suspension bridge as the engineering background， the influence of the aerodynamic contour of a Π-shaped stiffening girder on the measured wind parameters at the main girder level was studied， and a correction method for the measured wind angle of attack （AOA） was proposed based on the field measurement and Computational Fluid Dynamics （CFD） method. Moreover， the field measurements of wind speeds and AOAs at different measurement points on the Π-shaped stiffening girder were carried out for five months. The distribution of wind parameters along the bridge axis was analyzed， and the wind speeds and the wind AOAs measured by the anemometers on the windward and leeward sides at the stiffening girder level were compared， respectively. Furthermore， the CFD method was utilized to simulate the flow field of the stationary girder to investigate the influence of incoming wind speeds and wind AOAs on the wind parameters measured by the anemometer installed at different positions of the Π-shaped stiffening girder. Additionally， a correction formula for the wind AOAs measured at the windward side of the Π-shaped stiffening girder was proposed based on the numerical simulation. The results show that the measured wind speeds are relatively close within the main span while those are lower at the side span. The wind AOAs measured on the windward side are significantly larger than those on the leeward side of the girder， while the wind speeds are almost the same on both sides. The numerical simulation results of the wind parameters measured on the windward and leeward sides are consistent with the measured results， and the flow around the girder has a specific influence on the wind AOAs within 20 m of the girder. Via the wind AOAs data measured on the windward side， the rational wind AOAs results can be obtained based on the correction method proposed in this study.

Abstract:Based on the method of UAV photogrammetry and random vibration test， this paper explores a method to evaluate bridge conditions under non-closed traffic conditions. Using a bridge in Changsha as an example， the bridge line shape， bridge disease， and dynamic characteristics were studied by using structural identification theory and UAV oblique photography technology. The UAV 3D oblique photography technology is used to reconstruct the 3D scene model of the target bridge and to extract the bridge line shape through the point cloud model. The UAV measurement line shape is compared with the leveling line shape and the design line shape. It is found that the average value of the line shape difference is 0.034 m， the average relative deflection value is L/12705， and the standard deviation is 0.035 m. The two results are close， which verifies the reliability of the UAV measurement bridge line shape and that there is no significant change in the appearance size during the operation of the bridge. The small UAV technology was used to investigate the appearance quality of the target bridge， and the pier disease and bearing damage characteristics were investigated. The results showed that the bridge had no structural damage. Through the environmental vibration test of the structure， the Complex Modal Indicator Function （CMIF） method is used to capture the structural modal parameters （frequency， mode shape， damping ratio） with the load of driving and wind as the excitation. By comparing the historical baseline modal data 10 years ago and considering the influence of ambient temperature， it is judged that the structural stiffness of the bridge does not decrease significantly. The research shows that the comprehensive evaluation technology of bridge health status using “UAV 3D modeling linear measurement+UAV disease investigation + random vibration test” is feasible， economical， and practical.

Abstract:To study the influence law of helices with different parameters on the Vortex-induced Vibration （VIV） characteristics and aerodynamic force of stay cables， the wind tunnel test of a section model was carried out for smooth cables and those with helices. The helix parameters include number， diameter， and winding spacing. The results show that the VIV amplitude of one helix wound on the surface of the stay cable is much larger than that of two or three helices. In the subcritical Reynolds number region， the mean drag coefficient of the stay cable with two helices is smaller than that of the smooth cable， which can be reduced by 10.9% at most. The mean drag coefficient of stay cable with two helices is smaller than that with three helices. When the stay cable is wound with two helices with s=8D， the VIV of the stay cable wound with d<15 mm is obvious. The helix with d≥15 mm can effectively reduce the amplitude of the VIV， and its fluctuating lift coefficient can be reduced obviously compared with the smooth cable， which proves the effectiveness of the winding large-diameter helices in controlling the VIV response. The influence of the winding spacing on the VIV of the stay cable is related to the diameter of the helices. When the diameter is small， the influence of the winding spacing on the VIV is not marked. When the diameter is large， spacing directly affects the suppression effect of VIV.

Abstract:Aiming at researching the effect of additional attack angle on the flutter performance of the triple-box girder， based on wind tunnel test of section model and theoretical calculation， additional attack angles under initial wind attack angles were obtained， and the critical flutter wind speed under the initial wind attack angle was calculated by Computational Fluid Dynamics （CFD） numerical simulation. Results indicated that the initial wind attack angle on the triple-box girder and transverse beams between the girders had a remarkable influence on the additional attack angle. Moreover， the section model encountered a significant additional attack angle effect when the initial wind attack angle was in the range of 0°~ +7°， and presented “hard flutter” since the torsional vibration amplitude increased rapidly with wind speed once flutter was triggered. It is difficult to directly obtain the precise critical flutter wind speed of the triple-box girder through a free vibration test on a section model. For suspension bridges with a main span of 3 300 m， the additional attack angle before flutter instability can reach above +7° with an initial wind attack angle of 0°， and lead to an obvious reduction of critical flutter speed. In other words， the aeroelastic stability cannot be fully utilized. Therefore， great attention should be paid to the additional attack angle effect of triple-box girders in the wind resistance design of super-long span bridges.

Abstract:To address the issue of excessive static displacement in traditional Tuned Mass Dampers （TMD） for low-frequency and long-span bridges， this study introduces the implementation of a Tuned Mass Damper Inerter （TMDI） to increase flutter critical wind speed in such structures. This paper presents a TMDI configuration specifically designed for bridge flutter control and， based on a two-dimensional coupled flutter theory， develops the motion differential equation for the bridge-TMDI system， subsequently deriving the coefficient polynomial characteristic equation. Utilizing the Routh–Hurwitz stability criterion， the flutter critical wind speed is determined. A simply supported beam with an ideal planar cross-section serves as a case study to evaluate the efficacy of TMDI in bridge flutter control and to explore the influence of various TMDI parameter settings on flutter control. The findings reveal that TMDI can effectively increase the bridge flutter critical wind speed. Compared with the traditional TMD， the integration of an inerter may have a minor impact on reducing the control effect； however， it can notably decrease the static displacement of the mass block， thus offering great practical value in real-world engineering applications.

Abstract:According to the electrochemical corrosion principle of zinc-aluminum alloy coating， combined with the data of coating microstructure test， uniformity test， and neutral salt spray corrosion test， a two-dimensional corrosion cellular automaton model of zinc-aluminum alloy coating was developed， and the corrosion process of zinc-aluminum alloy coating was simulated. Then， the influence of the coating metal oxidation， the coating metal hydration， the hydrate dissolution， and the coating non-uniformity on the corrosion process were analyzed. The research results indicate that the two-dimensional corrosion cellular automata model can accurately simulate the corrosion process of zinc-aluminum alloy coatings. During the coating corrosion， metal oxidation plays a leading role， and metal hydration and metal hydrate dissolution are secondary factors. At the same time， the oxidation probability is nonlinearly proportional to the amount of metal corrosion per unit area， the dissolution probability is nonlinearly proportional to the amount of metal corrosion per unit area， and the hydration probability is inversely proportional to the amount of metal corrosion per unit area. The uneven distribution of metal content in the coating has little effect on the coating corrosion process， but the coating uneven thickness leads to premature corrosion of the iron matrix in local areas， resulting in the appearance of pitting pits.

Abstract:Two ultra-high performance concrete （UHPC） composite slabs with diamond-shaped joints （JF-L）and inverted T-shaped joints （JF-T） were designed and made to improve the wet joint of the UHPC-NC composite slab bridge in the coastal splash zone. An experimental study was carried out on these two specimens， and the interfacial stress was analyzed. The stress response of the entire progress from the initial crack to the failure of the composite slabs is obtained. Meanwhile， based on the experimental results and previous studies， the calculation formula for the interface cracking load of the UHPC-UHPC joint was developed. The results show that two specimens show bending-shear failure. The reinforcement ruptures at the reinforcement ratio changing section of the shear span due to the closer placement of reinforcement in the joint interface. The crack resistance and ultimate failure loads of the two specimens are nearly the same. The crack resistance is dominated by the interface behavior of the joint. Different configurations of wet joints have few impacts on the bending behavior of specimens. The nominal cracking stress of JF-L and JF-T is 4.71 MPa and 4.74 MPa， respectively， meeting the serviceability requirements of the coastal bridge. JF-L shows more cracks and lower cracking speed than JF-T. The UHPC-NC interface of JF-T is sound during the test， indicating great ductility and crack resistance of JF-T. The calculation formula for the cracking load of the UHPC wet joint was proposed and validated by experimental results in this and previous studies. In practical projects， the diamond-shaped UHPC joint is recommended as easy to prefabricate and construct.

Abstract:The drag coefficients of Square Towers with Circular Members （STCM） in Reynolds number range from 307 to 906 are performed by wind tunnel test with section model. The effects of the solidity ratio， turbulence intensity， and wind speed on the drag coefficients are analyzed. Based on the experimental results， the applicability of three Reynolds Averaging Navier-Stokes （RANS） models in Computational Fluid Dynamics （CFD） for the aerodynamic force simulation of STCM is compared. The variation law and influencing factors of drag coefficients in the range of Re=300~2.5×104 are further analyzed. The experimental results show that the mean drag coefficient of STCM is maximum at 45° and minimum at 0° and 90°. The mean drag coefficient decreases with the increase in solidity ratio and wind speed and increases with the turbulence intensity. The fluctuating drag coefficient is not affected by the solidity ratio and wind direction angle but increases with the wind speed and turbulence intensity. The CFD results show that the SST k-ω model is more suitable for calculating the mean drag coefficient of STCM at Re≤2.5×104. With the increase of Reynolds number， the mean drag coefficient of the STCM decreases sharply at Re= 300~585， slightly slows down at Re=585~1 090， and remains basically unchanged at Re=1 090~2.5×104. Compared with the drag coefficient of upstream members， the values of downstream members completely immersing in the wake of upstream members can be reduced by about 50%.

Abstract:To study the mechanical behavior of Cold-formed Thin-walled Steel （CFS） C-shaped compression members strengthened by batten plates， the nonlinear finite element model of C-shaped steel was established by using ANSYS software. The correctness of the modeling method was verified by comparing the test failure characteristics and ultimate bearing capacities. The effects of batten spacing and eccentricity on the ultimate bearing capacity and buckling mode of batten-strengthened CFS C-section members under axial and one-directional eccentric compression were studied. The results showed that， with the decrease of batten plate spacing， the buckling deformation of C-shaped members changed gradually from distortion buckling to local buckling， and the ultimate bearing capacity of axial compression and eccentric compression members increased gradually. With the increase of eccentricity， the ultimate bearing capacity of eccentric compression members decreased gradually. It is suggested that the batten spacing of CFS C-shaped steel axial compression and eccentric compression members strengthened by batten plates should be taken as λ/3 and λ/4， respectively. Based on the mechanical characteristics and direct strength method of CFS C-shaped steel members strengthened by batten plates， a modified formula for the ultimate compressive capacity of CFS C-shaped members strengthened by batten plates was proposed， and the accuracy of the formula was verified.

Abstract:To study the local buckling behaviour and failure mechanism of stainless steel components in fire， axial compression tests of eight austenitic welded I-section stub columns were introduced， which included six elevated temperature tests and two room temperature reference tests. In addition， this study analyzed their instability mode and local buckling bearing capacity， meanwhile， the important influence factors on column local stability were analyzed using numerical parametric studies. The current European design method was evaluated on the basis of the experimental and numerical data. The results show that as the temperature increases， the stiffness of the specimen decreases， the buckling deformation increases and the local buckling resistance decreases； the initial imperfection has little effect on the resistance of the specimen， and the section width-thickness ratio is the key factor affecting the local buckling resistance； the predictions of the current Eurocode design method are relatively accurate.

Abstract:The flexural tests of 6 sets of Q460 tube-gusset joints， involving three parameters of reinforced ring plates （unreinforced）， gusset plate widths， and axial compression force of the tube， were carried out. The results show that the load-bearing capacity of the unreinforced specimens is lower than that of the specimens with reinforced ring plates because of the stress concentration in both ends of the gusset plate. With the ratio of the width of the gusset plate to the diameter of the tube increasing from 2.5 to 3.0， the load-bearing capacity of the specimen with unreinforced， 1/4， and 1/2 ring plates is increased by 6.10%~16.07%， 13.36%~20.68%， and 9.61%~12.34%， respectively. Compared with the unreinforced specimens， the load-bearing capacity of the specimen with 1/4 and 1/2 reinforced ring plate is increased by 102.86%~130.73%， and 129.88%~166.33%， respectively. With the axial compression ratio of the tube increasing from 0.10 to 0.23， the load-bearing capacity of the specimen of unreinforced， 1/4， and 1/2 reinforced ring plate is decreased by 10.05%~17.77%、10.79%~16.20%、4.74%~7.05%， respectively. Several design guidelines are too conservative for the joint without reinforced ring plates. The load-bearing capacity calculation method of the Q460 tube-gusset joint was proposed considering the effect of the reinforced ring plate and the axial compression force of the tube， which agrees well with the test results.

Abstract:Aiming at the connection reliability problem of precast steel pipe-concrete composite columns， a post-stressed strengthened core-pipe connection is proposed. The columns are connected by Q420 high-strength core pipes， and anchored into the pre-embedded steel pipe of the column. The connection is strengthened by the post-tensioned unbonded tendons through the overall height of the structure. To investigate the seismic performance of the precast composite column， a low-cycled horizontal load test was carried out on a cast-in-place comparison specimen and three precast specimens with different core pipes. The seismic performance of each specimen was comprehensively evaluated by comparing the failure mode， hysteretic curve， skeleton curve， bearing capacity， stiffness， displacement ductility， and energy dissipation. Based on the test results， finite element parameter analysis of core pipe diameter and thickness， core pipe length， connecting cover plate thickness， and prestressed stress was carried out. The test and parameter analysis results show that， precast specimens can generally achieve equivalent cast-in-place seismic performance， but the adverse effects of increased structural seismic force due to increased stiffness should be considered. With the increase of the cross-sectional specification of the core pipe， bearing capacity， stiffness， and ductility of the column gradually increase， and the precast specimen with D152 mm×16 mm core pipe is recommended. Cross-section with a larger moment of inertia and equivalent area is proposed for core pipes. The bonding length of core pipes can be designed as 3 times the diameter. The thickness of the connecting cover plate can be determined by configuration. For the tested specimens， the prestressed stress is recommended to be determined within the range of 0.3fpyk~0.6fpyk.

Abstract:To further study the shear capacity of double steel plate shear walls with internal stiffeners connected at two sides， the finite element analysis of 36 double steel plate shear walls with different width-depth ratios and different height thickness ratios， which are internally stiffened with two-side connections， is carried out by using ABAQUS program. The analysis results showed that the use of thick plates or plates with a larger width-depth ratio can make the steel reach a higher average stress. The ratio of peak mean shear stress to yield mean shear stress has no obvious change rule with height-thickness ratio or width-depth ratio. The variation of strain-hardening ratio with height thickness ratio is between 1.17 and 1.21， and the variation with width-depth ratio is between 1.16 and 1.21， with an average value of 1.19. The formula for calculating the yield mean shear stress of double steel plate shear walls with internal stiffeners connected at both sides is given. The theoretical calculation formula derived from the displacement calculation principle of structural mechanics can accurately calculate the initial stiffness and initial equivalent shear modulus of an internally stiffened double steel plate shear wall. In addition， the equivalent cross brace model of double steel plate shear walls with internal stiffeners connected at both sides is established. It is verified that the equivalent cross brace model can accurately simulate the yield mean shear stress and initial stiffness of double steel plate shear walls with internal stiffeners connected at both sides with different width-depth ratios and different height-thickness ratios. It is shown that the design and analysis of double steel plate shear walls with internal stiffeners connected at both sides can be simplified by using the equivalent cross brace model.

Abstract:The appearance of ancient brick masonry structures strengthened along the horizontal mortar joint with the random distribution short fiber reinforced Engineered Cementitious Composite （ECC） with ultra-high toughness is almost unchanged， which is consistent with the principles of repairing old heritage buildings as the old one. With different fiber types such as polyethylene fiber （PE）， polyvinyl alcohol fiber （PVA）， and hybrid fiber （PVA and PE） reinforced ECC， the content of PE fiber， caulking spacing， and single-sided or double-sided reinforcement as variables， the diagonal shear test of brick masonry under shear compression was carried out to explore the shear and the cooperative performance of brick masonry reinforced with caulking ECC. The failure pattern， load-displacement curve， shear strength， and ductility are comparatively analyzed. The test results show that caulking ECC slows down the development of internal cracks and improves the shear-bearing capacity and the integrity of brick masonry. The maximum shear strength of the reinforced specimen is increased by 44.5%， and the maximum shear stiffness of the reinforced surface is about 5 times that of the unreinforced surface.

Abstract:To solve the problem of soil heat accumulation in the heat exchange process of energy piles， this research combined the advantages of phase change concrete and traditional energy piles， and the phase change material （steel ball） is mixed into concrete to form a phase change energy pile （PCEP） to improve the heat exchange efficiency and heat storage capacity of the energy pile. Then， the thermodynamic characteristics model test and simulation of PCEP in saturated sand were carried out， and the parameters， including the pile temperature and the temperature of the soil around the pile， axial strain in the pile， and pile top displacement under cyclic temperature load， were measured. The results show that the strain in the PCEP is lower than that of the ordinary pile after heating up， and the strain in the PCEP is close to that of the ordinary pile when the temperature is stable. The temperature lag effect and thermal stability of the PCEP are better， and the heat transfer efficiency is higher than that of ordinary piles， which helps alleviate the problem of soil thermal accumulation. The long-term cycle temperature of the pile is too high （50 ℃ during the day and 40 ℃ at night）， and the temperature in the pile exceeds the phase change temperature of the phase change material， so the energy storage effect becomes poor. An intermittent working environment is needed to achieve the best behavior in practice.

Abstract:To study the shear performance of concrete columns corroded by saline soil， apply different preloads to 11 polypropylene fiber lithium slag concrete columns and saline soil environmental erosion tests were carried out at the ages of 0 days， 90 days and 180 days. Then the specimens after erosion were subjected to low cycle level repeated loading tests. The influence of erosion time， axial pressure ratio， and preloads on the shear resistance of fiber lithium slag concrete columns and their failure mechanism were analyzed. The results show that the addition of polypropylene fiber and lithium slag can improve the shear resistance of concrete in saline soil environment； at the same axial compression ratio， with the increase of erosion time， the shear capacity of fiber-reinforced concrete columns shows a trend of first increasing and then decreasing； with the increase of the preloads， the shear capacity of the column increases first and then decreases. When the erosion time is fixed， the shear capacity of eroded members increases after setting the axial compression ratio from 0.2 to 0.3. According to the experimental results， and considering the influence of the fiber on the shear bearing capacity， a formula for calculating the shear bearing capacity of fiber columns is established based on the truss arch model， the fiber effect is considered as a tension field and is analyzed together with the stirrup. The average value of the ratio between the theoretical calculation value and the test value is 0.970， the standard deviation is 0.086， and the coefficient of variation is 0.090. Both two are better matched.

Abstract:To study the shear behavior of Ultra-high Performance Seawater Sea-sand Concrete （UHPSSC） beams reinforced with Fiber Reinforced Polymer （FRP） bars， the three-point bending loading test was used to explore the influence of shear span ratio， stirrup spacing and reinforcement type （CFRP， BFRP， HFRP） on the shear performance of FRP-UHPSSC beams. The test results showed that the cracks of FRP-UHPSSC beams developed rapidly during loading， and the beams had marked brittle failure characteristics. The mid-span load-deflection curves showed bilinear change before and after the first crack appeared. After the crack appeared， the slope of load-deflection curves became smaller， and the deflection growth speed increased. The shear capacity of FRP-UHPSSC beams decreased with the increase in shear span ratio and stirrup spacing. The cracking loads of test beams with the same shear span ratio were basically the same. The larger the elastic modulus of the reinforcement， the higher the shear capacity of test beams and the more vertical deformation is inhibited. Finally， the shear formulas of four codes are used for analyzing the test results. The minimum errors of the formulas for calculating the shear capacity of BFRP-UHPSSC beams and HFRP-UHPSSC beams are over 40% and 30%， respectively. The difference in the calculated shear capacity under the testing conditions of different reinforcement types varies greatly.

Abstract:This paper presents an Automated Crack Detection and Measurement （ACDM） method， which is based on the strain field obtained by the Digital Image Correlation （DIC） method and uses machine vision recognition to quickly obtain the distribution of cracks on the surface of the specimen in the corresponding damage state. Through three-point bending test， the Crack Mouth Opening Displacement （CMOD） value is obtained by the clip gauge and ACDM method， respectively. The CMOD was measured by the clip gauge as a reference. By analyzing the accuracy of this method， the results show that when CMOD was less than 0.05 mm， the identification error is less than 0.01 mm， and ACDM can be used to analyze the micro-crack concrete damage. Based on the compression tests of three groups of concrete cube specimens under different damage stages， the damage indexes obtained by ACDM method have corresponding numerical changes greater than 10-3， which is more sensitive and accurate than the current quantitative Cv-value of strain field.

Abstract:In view of the fact that the current tunnel monitoring measurement method cannot measure the displacement caused by the early excavation disturbance of the tunnel and the displacement lost by the failure to install the monitoring equipment in time， the deformation of the surrounding rock of the tunnel cannot be accurately predicted. Based on the full displacement theory， this paper proposes a relatively accurate calculation method for the loss displacement， the advance displacement， and the full displacement of the surrounding rock during tunnel excavation. Through the regression analysis of field monitoring data and the simulation of tunnel excavation by FLAC3D finite difference software， the research results show that： in the whole process of tunnel excavation， the internal displacement curve of surrounding rock is S-shaped including three stages of initial growth， rapid growth， and stable convergence. The influence range of the spatial constraint effect of the tunnel face is roughly within the range of four times the hole diameter. In the numerical simulation， the ratio of advance displacement to full displacement is the advance displacement ratio of surrounding rock， and the calculation result is 0.3~0.4， which is consistent with the actual deformation law of surrounding rock in the engineering site. Therefore， the full displacement of the internal deformation of the surrounding rock obtained by the research is more in line with the actual situation of the site. The research conclusions can provide a theoretical basis and reference value for the calculation of the full displacement of the internal deformation of the surrounding rock during the subsequent tunnel excavation process and the setting of the reserved deformation.

Abstract:To investigate the seismic reliability of transmission towers under the influence of Soil Structure Interactions （SSI） and tower-line coupling effect， a seismic reliability analysis method of transmission tower-line coupling system considering the SSI effect was proposed. Firstly， a simplified mechanical model of the transmission tower-line coupling system considering the SSI effect was developed， and the dynamic equation under stochastic seismic excitation was derived. Secondly， the random function-spectrum representation model was used to generate random seismic acceleration time-history samples， and the seismic dynamic response of the coupling system was analyzed. Thirdly， the maximum entropy and fractional moments based on GF-discrepancy points were used to predict the distribution function of the extreme-value variable of random seismic response， and the coupling system’s seismic reliability was solved. Finally， the effectiveness of the proposed method was verified through a transmission tower-line system from an actual project. The results show that： When the foundation soil category is type Ⅳ， after considering the SSI effect， the maximum vertex displacement and acceleration increase by 46.59% and 17.43%， respectively， and the failure probability increases by 70.32%. When considering the tower-line coupling effect， the maximum vertex displacement is reduced by 0.51%， the maximum vertex acceleration increases by 1.74%， and the failure probability is reduced by 15.71%； For the transmission tower-line coupling system considering the SSI effect， the foundation soil becomes softer， and the maximum vertex displacement and acceleration become larger. Compared with results using Monte Carlo Simulation （MCS）， the maximum relative error of the proposed method is 9.97%. The proposed method has high accuracy and efficiency.

Abstract:The wind pressure distribution law of the roof trough concentrator group is studied by large eddy simulation， and the numerical simulation results were compared with the wind tunnel experimental results to verify the accuracy of the large eddy simulation results. Then， the wind pressure distribution of a mirror group of nine flat roof concentrators was simulated under 0° elevation angle working conditions， and the most disadvantaged position of the flat roof concentrator group under different wind angles and the distribution law of its wind pressure coefficient were obtained. The results show that in the positive wind direction， the row of mirrors on the roof near the upstream of the incoming wind carries most of the wind load， and the wind pressure coefficient of the middle and rear rows of mirrors is 8.75% of that of the front row； in the oblique wind direction， the maximum wind pressure coefficient of the rear row of mirrors is 17.6% of that of the front row of mirrors； the corner position of the roof is the most disadvantaged position for the group of mirrors. The obtained conclusions can provide a theoretical basis for the design of the wind resistance of the group structure of slotted concentrators on flat roofs.

Abstract:To achieve the accurate and efficient classification of IFC components， an improved Multi-view Convolutional Neural Network （MVCNN） model is proposed， which introduces a self-attention module and a Long Short-term Memory （LSTM）network. To address the limitations of MVCNN model feature fusion， an LSTM_ATT module is designed. By adaptively adjusting the feature relationships of each view data and fusing the input data of each view according to the training attention weights， a more discriminative 3D shape descriptor is obtained. In turn， this improves the classification and detection performance of the model for similar IFC components. Finally， the improved MVCNN model is experimentally compared with the MVCNN using the IFCNet database， which comprises 20 classes of IFC components commonly used in the construction industry. The experimental results show that the overall accuracy of the proposed model for classification and recognition as well as the F1 value reach 88.27% and 86.72%， respectively， which is an improvement of 9.46% compared with the classification accuracy before the improvement， and the classification recognition between similar components is obvious.

Abstract:In a pursuit to develop strategies to amplify the resilience of freeway networks， this paper introduces an evaluation method of road resilience based on the combined weighting-cloud model. First， four topological structure indicators were selected， namely structure entropy， edge betweenness， freeway network density，and clustering coefficient， as well as two traffic status indicators： the travel time index and the traffic heterogeneity index. The resilience of the freeway network was comprehensively evaluated based on the topological structure and traffic status indicators. Then， the resilience of the freeway network was graded， the boundary values of evaluation indicators at different resilience levels were determined， and the characteristic values and certainty of the cloud parameters were estimated based on the backward cloud generator. Afterward， the indicators were weighted by combining the analytic hierarchy process and the entropy method. The membership degrees of different resilience levels were determined by calculating the weighted average， and the resilience level of the freeway network was detected according to the maximum membership degree. Finally， a case study was made for a freeway network to compare the combined weighting-cloud model method proposed in this study with the comprehensive fuzzy method. It is indicated from the research that the evaluation results of the two methods are similar. In contrast， the combined weighting-cloud model method reflects the actual status of the freeway network more objectively because it is free from the defect of randomness， which is included in the latter method.