Under long-term loading and environmental actions, the joints of prefabricated concrete bridges are prone to occur damage, which alters the transverse load distribution characteristics and may cause potential safety hazards such as single-plate load bearing. To investigate the influence mechanism of joint damage on structural mechanical performance, a multi-parameter damage simulation method based on interface slip normal stress and shear stiffness parameters (Knn, Kss, Ktt) was proposed. A refined ABAQUS finite element model was established, and the joint stress mechanisms of hollow slab girders, small box girders, and T-girders under single hinge joint damage and multi-hinge joint damage conditions were analyzed from three dimensions such as longitudinal normal stress, transverse normal stress and vertical shear stress. The research results show that the longitudinal normal stress at the joint is symmetric about mid-span, and damage mainly affects the stress distribution near the mid-span and supports. The transverse normal stress fluctuates near mid-span, and the more severe the damage, the larger the stress deviations are. The vertical shear stress exhibits an antisymmetric distribution, and the variation trend of stress curves after damage shows significant differences among different girder types. The overall stress distribution trend at the joint does not change markedly before and after damage. For hollow slab girders, the controlling order is transverse normal stress, vertical shear stress, and longitudinal normal stress; for small box girders and T-girders, the controlling order is vertical shear stress, transverse normal stress and longitudinal normal stress.

A new type of earthquake resilient precast hybrid pier was developed. The horizontal low cycle reciprocating load tests were carried out on the proposed hybrid pier and the RC pier counterpart. The test results show that the damage of the hybrid pier is mainly concentrated on the replaceable thin-walled steel tube, and the pier body remains elastic. At the same displacement angle, the residual displacement of the hybrid pier is significantly lower than that of the RC pier. When the displacement angle reaches the peak load of 2.0%, the residual displacement angle of the hybrid pier is only 0.2%, fully verifying its good self-centering ability. The ultimate displacement angle (θu) of the hybrid pier is about 5.6%, and its deformation capacity is much greater than that of the RC pier (θu =3.1%). Due to the better deformation capacity of the hybrid pier, its final cumulative energy consumption is greater than that of the RC pier. The flexural capacity of the hybrid pier is provided by the vertical axial force and the replaceable thin-walled steel tube. Based on this, the calculation formula of the flexural capacity of the hybrid pier is proposed.

Aiming at the problem of support elevation adjustment during incremental launching of steel truss girders with complex vertical curves, an optimization method based on the improved MOPSO algorithm was proposed. Firstly, a multi-objective optimization model for support elevation adjustment in steel truss girder incremental launching was established, with the maximum Mises stress in the main girder and the number of support elevation adjustments as optimization objectives, and the stresses of the launching nose and main girder within allowable limits as constraints. Secondly, the basic MOPSO algorithm was improved by dynamically adjusting the inertia weight and learning factors, introducing particle Gaussian mutation, and implementing an external archive maintenance strategy based on crowding; and the performance of the improved algorithm was verified. Finally, the proposed optimization method was applied to the incremental launching of the Luoxizhou Bridge in Ganzhou to validate its applicability. The research results show that the diversity and convergence of the Pareto solution set obtained by the improved MOPSO algorithm are significantly better than those of the basic algorithm. With the optimized method, the maximum Mises stress of the main girder during launching decreases from 118.762 MPa in the original scheme to 102.650 MPa (a reduction of 13.6%), and the total number of adjustments is reduced from 36 to 20 (a decrease of 44.4%). The proposed optimization method effectively reduces the number of support elevation adjustments, improves the structural stress state, and enhances the efficiency of incremental launching construction.

With the increasing development intensity of urban subway overpasses, the structural safety of shield tunnels is becoming increasingly affected by adjacent pile foundations. Based on a practical engineering case, a method for calculating the vertical displacement of a Pasternak-Timoshenko beam coupling model under pile foundation loading was proposed. First, the additional vertical stress generated at the tunnel axis by pile-tip and pile-side loads was calculated by the Mindlin solution. Subsequently, the shield tunnel segments were equivalent to Timoshenko beams placed on a two parameter Pasternak foundation, and the analytical solution for vertical displacement was derived through Fourier series analysis. Parametric analyses on the beam length (30.0-60.0 m and infinite length) and load distribution width (6.0 m, 12.0 m) were carried out. And a three-dimensional finite element numerical model was used to compare and verify the theoretical model, based on which, the influence of beam length and load distribution width on tunnel settlement was further analyzed. The results show that the proposed theoretical model for calculating vertical displacement is reasonable. When the beam length is 60.0 m and the load width is 12.0 m, the ERMS and EMA between the analytical and numerical solutions are 0.474 mm and 0.471 mm, respectively. However, the analytical solution still underestimates the peak settlement by 0.560 mm and the influence width by 0.1D~ 0.2D.

High polymer grouting technology for seepage control and reinforcement has been widely applied to the leakage and settlement repair of underground concrete structures in recent years. The flow and diffusion rule of self-expanding high polymer grouting material in concrete cracks is crucial for determining grouting parameters. However, the diffusion and solidification processes of high polymers in cracks with different roughness levels have not yet been clearly defined. Based on this, a self-expanding high polymer diffusion model considering the time varying viscosity characteristics and crack roughness of high polymers was established by the phase-field method. The research results show that crack roughness has a significant effect on polymer diffusion and stress response. When Z2≥ 0.195, crack stress and deformation increase sharply, indicating that Z2=0.195 can be regarded as the critical threshold for the transition of high polymer flow diffusion and crack mechanical response. Below this threshold, slurry diffusion is close to that of the parallel-plate model; above it, stress concentration and local damage are significantly intensified. At a grouting velocity of 5 mm/s, the maximum displacement of the rough crack with Z2=0.413 is 48.7% higher than that of the smooth crack, and the fluid pressure reaches 1.07×105 Pa. Comparisons among different crack profiles show that the larger the Z2, the earlier and faster the damage occurs and spreads. Compared with L3, L4 occurs damage 0.7 s earlier and reaches complete destruction in approximately 2.63 seconds, with a stress increase of 226.2%.

Focusing on the joint area of concrete anti-scour walls — a critical weak zone under chloride erosion environment, the corrosion characteristics and evolution patterns of steel reinforcement in the joint region were systematically investigated through combined experimental and numerical simulation approaches. Concrete joint specimens were designed and fabricated, and immersion and accelerated electrified corrosion methods were employed to measure electrochemical parameters such as corrosion potential and corrosion current density. The influence of chloride concentration and interface cracking on reinforcement corrosion was focused and explored, and numerical simulation was conducted to analyze the chloride diffusion behavior in the joint zone. The research results indicate that the risk of reinforcement corrosion in the joint area is significant. Chloride ions penetrate laterally along the joint interface and subsequently propagate longitudinally along the steel-concrete interface, forming a “lateral penetration-longitudinal propagation” dual-pathway erosion pattern. Increasing the chloride concentration from 5% to 10% leads to an average increase of 17% in the corrosion current density of reinforcement in the joint zone. The 0.2 mm interface crack not only advances the corrosion time of reinforcement corrosion but also increases the current density by 36%, and there is a significant two-dimensional diffusion effect of chloride ions in the crack area. The research results provide a quantitative basis for the durability design and maintenance of the joint area of concrete anti-scour walls.

The reducing environment at the sediment-water interface is a critical area for element transformation. The interface equilibrium of phosphorus element directly regulates the eutrophication process of water bodies, which is deeply influenced by subtle fluctuations in interface redox conditions. Taking the typical tributary of the Tuojiang River, the Laixi River, as the object, surface sediment samples were collected and subjected to laboratory-scale simulated flooding experiments. Combined with the Hedley sequential extraction method, the dynamic changes of different phosphorus speciation in reducing environments were systematically analyzed. The mechanism how key parameters such as overlying water pH, oxidation-reduction potential (Eh), Fe2+, total iron and dissolved organic carbon (DOC) influenced phosphorus release process of sediment was mainly explored. Results indicate that sediments in the Dazu section of the Laixi River exhibit weak alkalinity, and the main form of phosphorus element is calcium-bound P (Ca-P), followed by iron/aluminium-bound P (Fe/Al-P). During inundation, the sediment Fe/Al-P decreases significantly (p<0.01) while overlying water P concentrations increase concurrently, indicating that it is an important source of potential active phosphorus. Correlation analysis indicates that overlying water Eh exhibits a significant negative correlation with phosphorus concentration in water (p<0.01). Meanwhile, Fe2+ and DOC in the overlying water are positively correlated with total P in overlying water, exerting a certain promoting effect on sediment P release.

To improve the signal quality of fiber Bragg grating (FBG) sensors in monitoring the strain and temperature fields within asphalt pavement structures, an FBG sensing signal denoising processing method combining Savitzky-Golay (SG), singular value decomposition (SVD) and waveform suppression was proposed. First, convolution smoothing of the signal was carried out to reduce unnecessary fluctuations. Next, singular value decomposition was carried out, and the main signal feature patterns based on the size of singular values were selected out, achieving data dimensionality reduction and noise filtering processing. Finally, the data features were strengthened and non-essential signal fluctuations were suppressed to enhance data stability. The proposed method was applied for noise reduction processing of simulated signals of pavement structure response and measured structural response signals of asphalt pavement. The peak signal-to-noise ratio (PSNR) of the field-measured signal increased from 115.60 dB to 121.48 dB, with effective signals enhanced and ineffective signals weakened. The proposed method can effectively improve the credibility of signals and more accurately monitor the changes in internal strain response of asphalt pavement structures under driving loads.

Aiming at the problem of complex coupling mechanism and limited comprehensive performance of the model in analyzing the factors affecting the severity of accidents, a model of multifactor coupled accident severity (MFCAS) on mountainous expressways was designed. MFCAS classified the severity of accidents based on the number of casualties and property losses as constraints and constructed a multifactor coupled feature space. A gradient boosting decision tree based on significant factor identification algorithm (GBDT-FSIA) was proposed. By using local outlier factor algorithm to remove abnormal samples and then using heuristic mean clustering algorithm to partition longitudinal slope labels, the significant influencing factor set was determined, and its coupled effect was quantified. Research results show that under typical longitudinal-slope heterogeneity conditions, the road longitudinal slope, driver age, driving experience and weather are significant factors affecting the severity of accidents on mountainous expressways. During rainy weather, drivers with less than 10 years of driving experience and over 50 years of age have the highest risk of accidents, drivers with over 20 years of driving experience and aged from 30 to 50 in high longitudinal slope environments exhibit significant risk suppression effects, and the combination of rainy days and high longitudinal slopes has a significant synergistic amplification effect. Compared with traditional models, the average accuracy of severity prediction of the proposed model can reach 85.2%, the average accident recall rate increases by 3.13%, and the AUC value increases by 2.04%.

Distracted driving has become a major contributing factor to traffic accidents. To investigate the temporal instability and unobserved heterogeneity of the factors influencing the severity of distracted driving accidents, the partial time constraint method was employed to address temporally unstable parameters and a random parameter Logit model that accounted for both mean and variance heterogeneity was developed. Thirty-seven potential influencing factors were extracted from aspects such as driver, road, vehicle and environment. Taking the severity of three categories of accidents as the dependent variables, multiple comparisons of log-likelihood values were conducted by capturing variations in the means and variances of the random parameters. The temporal instability of different influencing factors was verified, and the causal mechanisms of accident severity based on marginal effects were analyzed. The results show that distracted driving accidents exhibit significant temporal instability. “Male gender, collision with fixed objects, and wet road surfaces” have random effects during 2017-2019; their means are significantly associated with variables such as SUVs, not wearing a seat belt and the elderly, and their variances are influenced by male gender and wet road surfaces. Factors including not wearing a seat belt, rollover accidents, trucks, and speed limits above 60 km/h significantly increase accident severity. The research results provide certain valuable references for improving traffic safety regulations and preventing driving accidents.

To address the issue of policy update lag in urban traffic signal control by the deep Q-network (DQN) algorithm, an improved double-layered exploration and multi-step reward algorithm, that is MS_DQN (multi-step deep Q-network) algorithm, was proposed. The proposed algorithm constructed a double-layered exploration mechanism by combining(ε-greedy strategy) and Boltzmann strategy, enabling the agent to explore unknown spaces more comprehensively as well as balance exploration and exploitation. Meanwhile, rewards from multiple future time steps were introduced. Through a multi-step reward mechanism, sample correlation was reduced, higher long-term rewards were obtained and the learning stability was enhanced. Experimental results on the simulation of urban mobility (SUMO) platform demonstrate that the MS_DQN algorithm can significantly improves intersection throughput efficiency, providing an effective solution for urban intelligent traffic control.

To address the problem that existing graph neural networks diluted critical features of long trajectories and failed to distinguish fine-grained category differences due to average or sum pooling in the graph readout phase, a graph isomorphism network fusing a global attention mechanism (GIN-GA) was proposed. The improvements of the proposed model included: retaining the Delaunay triangulation spatiotemporal graph composition strategy to preserve topological integrity; introducing global attention pooling in the feature aggregation layer to adaptively assign trajectory point weights, and enabling the model to focus on the most discriminative key nodes for classification; employing a weighted cross-entropy loss to address the class imbalance problem in AIS data. Research results show that GIN-GA achieves an F1 score of 92.81%, which is 3.32%, 2.92%, 3.01%, and 2.29% higher than those of the original GIN, mainstream graph convolutional networks (GCN), graph attention networks (GAT), and graph sample and aggregate networks (GraphSAGE), respectively. The recall rates of two minority samples, passenger ships and tugboats, are improved by 4.70% and 7.00%, respectively. The proposed model effectively enhances ship recognition accuracy and provides a reference for improving the intelligence level of maritime supervision.

The running tires of straddle type monorail vehicles have the phenomenon of interface delamination during operation. To effectively enhance the safety and service life of the running tires and reduce the operation and maintenance costs of the running tires, the running tires of straddle type monorail vehicles were taken as the research object, and a finite element model of the running tires was established. Through the stress and strain results under the thermal coupling effects of three working conditions (steady-state rolling, side bias, and side inclination), the Lake-Lindley model was combined with the maximum tear energy method to fit the fatigue parameters, and the thermal fatigue life of the running tires under multiple working conditions was simulated by the FE-safe software. Taking the fatigue life of the running tires as the target, the angle and spacing of the belt layer cord were the optimization variables. Based on the response surface method and genetic immune algorithm, the research on optimization of the running tire belt layer cord parameters was carried out. The research results show that the tire tread grooves and the end of the belt layer in the steady-state rolling condition are high-risk areas, and the load increase significantly shortens the lifespan. Fatigue damage occurs at the tire body rubber rebound point under the side bias condition and at the tire surface side inclination end and triangular rubber area under the side inclination condition. By optimizing the angle and spacing of the belt layer cord of the running tires, the fatigue life of the running tires has increased by 24.4%. The research results provide a certain theoretical basis and methodological support for the fatigue optimization design of the running tire structure of straddle type monorail vehicle.

Fogging on vehicle windshields can severely obstruct the driver’s field of vision and endanger driving safety. Traditional automatic defogging systems rely on temperature and humidity sensors to determine fogging risk, making it difficult to accurately perceive the complex and non-uniform microclimate distribution inside the vehicle, which easily causes delayed defogging response or control redundancy. To address this issue, a lightweight transmission estimation method was proposed from a visual perception perspective. The pixel-level transmission distribution on the windshield surface was directly estimated from a single image to characterize the fogging state. Aiming at the spatial non-uniformity of fog distribution and the directional nature of its physical diffusion, a lightweight network named MSF-Net was designed, which integrated a Manhattan self-attention mechanism with a frequency-domain dynamic aggregation module. Specifically, the Manhattan self-attention mechanism models had horizontal and vertical feature dependencies, incorporating physical diffusion path prior to enhance the ability to characterize fog structures. The frequency-domain dynamic aggregation module adaptively decoupled background texture and fog features in the frequency domain, effectively mitigating the interference of high-frequency background details on transmission estimation. Furthermore, to overcome the scarcity of real-world fog data, a physics-optics coupled data synthesis method based on COMSOL multi-physics field simulation was proposed, constructing the synthetic windshield fog image with physical consistency and the corresponding transmittance true values. Research results demonstrate that MSF-Net outperforms mainstream methods such as DCP and DehazeNet in terms of PSNR and SSIM on the synthetic dataset, meanwhile it also exhibits significant lightweight advantages. The proposed model has been successfully deployed on a Raspberry Pi 5 embedded platform, and its inference speed can meet the application requirements of periodic state updates and closed-loop regulation in onboard defogging control systems, thereby providing a viable visual perception solution for intelligent defogging systems.