The severity of freeway traffic accidents is influenced by a multitude of factors, among which short-term traffic flow characteristics prior to an accident play a particularly critical role. To systematically analyze the impact of short-term traffic flow conditions on accident severity, this study draws on historical traffic accident data, ETC gantry passage data, and meteorological data from three expressways in Guangdong Province—Nan Erhuan Expressway, Jingguang Expressway, and Western Coastal Expressway—for the years 2021–2022. A random parameter Logit model incorporating mean heterogeneity is constructed to explore the heterogeneous characteristics of accident - related factors. This research identifies 29 potential variables from four perspectives: road characteristics, environmental characteristics, traffic flow characteristics, and accident characteristics. Then, it models accident severity using a standard multinomial Logit model, a random parameter Logit model, and a random parameter Logit model with mean heterogeneity. By comparing the goodness of fit of the models using the pseudo R-squared, Akaike Information Criterion, and Bayesian Information Criterion, the results demonstrate that the random parameter Logit model with mean heterogeneity outperforms the others, capturing the heterogeneous characteristics of accident - related factors more accurately. Further evaluation of the impact of different factors on accident severity based on the average elasticity coefficients of the variables shows that, at a 99% confidence level, 22 parameter variables related to road, environmental, accident, and traffic flow characteristics significantly affect accident severity. Specifically, certain factors such as a six-lane design and increased visibility reduce accident severity, while others, including road rescue processing time, average speed of large vehicles, proportion of large vehicles, and speed difference between large and small vehicles, intensify accident severity when increased. The findings of this study provide a scientific basis for freeway accident prevention and management.

Understanding the traffic accident mechanism of autonomous vehicle is an important prerequisite for effective prevention and control of safety risks. The analysis of the key causes of autonomous vehicle accidents is usually based on small samples and imbalanced data, but such models have low prediction accuracy for minority classes. The analysis framework based on data augmentation can improve the prediction accuracy of the model for the minority class. Conditional GAN for Generating Synthetic Tabular Data(CTGAN), Copula Flows for Synthetic Data Generation(CopulaGAN), the Synthetic Minority Over-sampling Technique (SMOTE) and the Adaptive Synthetic technique (ADASYN) increased the sample size and balanced the data set. Then, the best machine learning classification algorithm is determined based on the synthetic data. Finally, combined with the SHAP framework, the importance of the key causes of accidents is quantified, which can accurately analyze the key causes of autonomous vehicle accidents. The results show that CTGAN can effectively solve the problem of poor classification performance of small sample size and imbalanced dataset. CTGAN combined with random forest classification algorithm for model training can significantly improve the prediction performance of the model for autonomous driving accidents. Road conditions, driving at night, and the degree of intersection and street transformation are the key causes of autonomous vehicle accidents. The research results can provide reference for the construction of autonomous driving vehicle test scenarios and the transformation of active road infrastructure.

Long-distance heavy truck drivers are prone to fatigue when driving on highways, which threatens driving safety. To explore the characteristics and patterns of fatigue in heavy truck drivers during continuous highway driving, an on-road driving experiment was conducted. The drivers' behavior and eye movement indicators were recorded using cameras and an eye-tracking device. The analysis was carried out through function fitting, driver behavior characteristic analysis, and the construction of a Multilayer Perceptron (MLP) model. The results indicate that the safe threshold for continuous driving time for truck drivers is between 3.6 and 3.7 hours. In terms of Passive Task-related Fatigue, based on average blink duration, drivers experience a sense of fatigue 5 to 15 minutes before they exhibit active fatigue intervention behaviors 48.69% of the time. The frequency of checking the rearview mirror and engaging in non-driving-related small actions correlates with average blink duration. Using these behaviors as substitutes for eye movement indicators, the detection rate of passive fatigue during the driving process can reach 80.2%. The findings can provide a reference for video monitoring and warning systems for heavy truck driver fatigue, and contribute to enhancing the safety of truck driving.

To study the spatial characteristics of the wind field at the bridge-tunnel connection section in complex mountainous areas, the mountainous terrain within 8km in diameter at the junction of G318 and S217 is selected as the research background. The mountain models of five lengths of bridge-tunnel connection section are established based on the digital elevation model. The spatial distribution characteristics of the wind field at the bridge-tunnel connection section under 16 kinds of flow conditions are obtained by using the numerical simulation method. The results show that, under the condition of the approximately constant mountain slope, affected by the actual complex terrain, there are some differences in the characteristics of transverse wind speed, vertical wind profile, and wind attack angle at the bridge-tunnel connection sections with different lengths, but the overall trend is similar. When the incoming flow is perpendicular to the bridge-tunnel connection section, the wind speed at the midspan is the maximum due to the canyon acceleration effect. This acceleration effect increases with the decrease of the bridge-tunnel connection section length. The incoming flow in other working conditions is reduced due to the reduction effect of high and steep mountains on both sides, and the wind speed is the minimum along the bridge-tunnel connection section. High and steep mountains and river bends affect the vertical transverse wind speed distribution. In the canyon with lower elevation, the shorter the length of the bridge-tunnel connection section, the more significant the impact. The characteristics of the wind angle of attack are also greatly affected by the terrain, and the overall performance is dominated by the negative angle of attack. The variation rules obtained from the numerical simulation of the wind field at the bridge-tunnel connection section in complex mountainous areas can provide some guidance and reference for studying traffic safety at the bridge-tunnel connection section.

Elucidating the interrelationship between pavement and tire factors and skid resistance is of significant importance for reducing traffic accident rates caused by insufficient pavement skid resistance. Three-dimensional point cloud data and corresponding coefficients of adhesion from 100 actual grooved cement concrete pavement test points were obtained through field standard testing, and after horizontal correction and noise reduction processing, inverse modeling is performed to establish a pavement model that retains road texture. Subsequently, a tire-pavement model is assembled in ABAQUS. Finally, the effectiveness of the model is verified by comparing the finite element back-analysis values with the experimental measurement values, and the impact of road texture, tire pressure, and speed on pavement skid resistance is analyzed. The results indicate that as the density of road surface peak points (S_pd) increases, the adhesion coefficient shows an upward trend at both high and low speeds, with a more significant influence at low speeds; with the rise of tire pressure, the adhesion coefficient at each test point exhibits a relatively consistent downward trend, with no significant difference in the magnitude of the decrease. Compared to speed, the impact of tire pressure on the adhesion coefficient is less significant; as speed increases, the road adhesion coefficient tends to a stable value determined by the macrotexture of the road surface, while the influence of speed itself on the adhesion coefficient is relatively limited.

The dynamic performance of the interface connection between old and new concrete in prefabricated concrete members is studied based on the separated Hopkinson pressure bar test method. The test parameters include the average depth and number of grooves, the tilt angle of the interface between new and old concrete, the combination of three different concrete strength levels and whether reinforcement is installed. The results show that the impact pressure of the test and the inclined angle of the interface of the specimen have significant effects on the failure mode of the specimen, and the specimens can be divided into five categories according to the failure mode. With the increase of impact pressure, the strain rate of specimens with different interface inclination angles increases gradually. With the increase of strain rate, the peak stress of each specimen increases. The peak stress of the specimen with the interface inclination of 40°, the depth of grooves, the number of grooves and the strength grade of new concrete is higher. The interfacial shear stress of each specimen increases with the increase of strain rate. The shear stress is greater at the interface of specimens with an interface inclination of 40°. The reinforcement can increase the peak stress of the specimen under certain conditions. The interface bearing capacity formulas for calculating the unslotted specimens under static loading, the slotted specimens under static loading, and those under dynamic loading were derived in specimens without reinforcement. The results show good agreement with the experimental outcomes.

As a typical discontinuous medium, the creep behavior of discrete granular materials strongly influences the formation and evolution of geohazards such as landslides and mudslides. However, systematic studies on the interaction mechanism between particle size and bias stress and its effect on creep behavior are still insufficient. To reveal the coupling mechanism of particle size and bias stress on creep behavior, this study carried out an indoor creep test of silica bead particles under multiple working conditions and systematically analyzed the influence of different particle sizes and bias stresses on the creep characteristics of granular materials. On the basis of the Derec creep model and experimental results, a quantitative calculation model of the creep state of granular materials was constructed, and the influence mechanism of the particle size on the creep parameters of the system was elucidated. The results show that the creep behavior of the granular system essentially reflects the dynamic balance between the internal deformation of the particles and their resistance to deformation. The creep parameter has a significant effect on the creep properties of the system by regulating the interparticle slip and creep behavior. Specifically, as the particle size increases, the creep value of the system increases significantly, and it becomes easier to enter the liquid-like flow state, whereas the deformation resistance of the system weakens and becomes more sensitive to bias stress. The particle size is positively correlated with the initial state parameters and the characteristic strain of the system, whereas it is negatively correlated with the viscous coefficient, the critical creep velocity and the critical creep stress.

Through direct shear and consolidation tests, the variation laws of cohesion c, internal friction angle φ, and compressibility index of granite residual soil (GRS) under drying-wetting cycles (DWC) and construction vibrations, as well as the parameter prediction formulas, are rationally obtained, which are then applied to the refined numerical simulation of deep excavation of a subway station. The experimental results indicate that c exhibits a nonlinear decay with the increase of the number of DWC and the vibration time, φ shows a fluctuating upward tendency with the increase of the number of DWC, and the effect of vibration on φ decreases as the number of DWC increases. The compression coefficient a_v1-2 increases linearly with the number of DWC without vibration and fluctuates after the vibration. The reference tangent modulus decreases slowly with the increase of number of DWC without vibration, but decreases sharply, then slowly, and finally increases after the vibration. By considering the effects of DWC and construction vibration, the foundation pit is divided into 6 impact zones, and the model parameters are selected from the deteriorated soil parameters corresponding to each impact zone. It is shown that the refined finite element simulation results agree well with the field monitoring values of the surface settlement and the horizontal displacement of diaphragm wall, which can provide a practical guidance for the design and construction of the residual soil foundation pit.

This study aims to improve the transmission stability of vehicle gearboxes during operation by focusing on the Ravigneaux planetary gear system. A comprehensive analysis of the nonlinear vibration characteristics of the Ravigneaux planetary gear transmission is conducted, and a dynamic model incorporating multiple coupled nonlinear factors is established. The model accounts for time-varying mesh stiffness, backlash, comprehensive transmission error, dynamic meshing force, and time-varying friction. Based on this, a set of nonlinear dynamic differential equations is derived, which are solved iteratively using the Runge-Kutta numerical integration method to obtain the system's dynamic response under varying external excitation frequencies. The system's complex nonlinear dynamic behavior is revealed through time history diagrams, spectrum diagram, phase diagram and Poincaré diagram. Furthermore, with other system parameters held constant, bifurcation diagrams and three-dimensional waterfall plots are employed to analyze the influence of excitation frequency on the vibration response. The results show that as the excitation frequency changes, the system’s vibration response evolves from chaotic states to periodic bifurcations, and eventually transitions to a stable single-period motion. This research provides a theoretical basis and engineering reference for adjusting excitation parameters to suppress non-steady-state vibrations and enhance the operational stability of the transmission system.

RV reducers have higher power density compared to traditional planetary and harmonic reducers, and are widely used in equipment operating in special and extreme working conditions and industrial robots. However in special and extreme environments, especially in high and low temperature alternating service environments, the components in RV reducers are supposed to have extremely high machining accuracy because of micrometer level errors can cause multi tooth meshing become multi tooth interference under loading conditions, which affects the lubrication of various components and ultimately leads to wear failure. However, there is currently a lack of systematic research on the wear behavior of RV reducers under special and extreme working conditions. According to the tribology and meshing theory, the motion and force analysis of each key component are first carried out. Then, a mixed lubrication model of the RV reducer components is established, and the cycloidal-pin wheel and needle bearing are confirmed as the insufficient lubrication parts in the RV reducer. Based on lubrication analysis, a wear model of the RV reducer is established, and finally, the wear model is calibrated based on equivalent experimental data. This model can predict the wear of RV reducers under special and extreme working conditions and provide theoretical support for improving their service performance.

The motion performance of humanoid robots has not fully reached the level of human beings, which is one of the factors affecting its large-scale industrial application. Such performance gaps stem not only from control algorithm limitations but also from mechanical design constraints, where the leg configuration plays a decisive role in the robot's dynamic stability, payload capacity, and energy efficiency.The leg configuration has an important effect on the motion performance of humanoid robot. The origin, history and status of leg configurations of humanoid robots at home and abroad are studied. Currently, the leg configurations of humanoid robots are primarily categorized into three types: serial, parallel, and hybrid serial-parallel configurations, with their structural characteristics significantly influencing motion performance.The series, parallel and series-parallel configurations are analyzed and their performance characteristics are compared. The serial configuration offers a large workspace and high flexibility, but its relatively lower stiffness—due to the extended joint transmission chain—compromises load-bearing capacity. In contrast, the parallel configuration provides high stiffness and rapid dynamic response, albeit with limited motion range. The hybrid serial-parallel design combines the strengths of both, achieving balanced stiffness and flexibility, which has increasingly made it a key research focus in recent years. The technical difficulties and hot spots in the study of leg configuration are discussed. The development trend is pointed out: the leg is developing from single series configuration to parallel and series-parallel configuration. The leg actuator develops from rigid actuator to elastic actuator and quasi direct drive actuator. The legs only use the rotary actuator to develop in the direction of rotation and linear actuator. The control is developing from position control to torque control and hybrid control of force and position. The research is of great significance for promoting the technological progress and industrial landing of humanoid robots.

This study determined the minimum inhibitory concentration (MIC) and minimum bacterial concentration (MBC) of plant derived active substances, and screened a synergistic combination of phloretin (Ph) and lauraldehyde (La) against the spores of Alicyclobacillus acidoterrestris by measuring the fractional inhibitory concentration index (FICI). The MIC of phloretin is 0.1 mg/mL, and MBC is more than 6.4 mg/mL; The MIC of lauraldehyde is 0.4 mg/mL, and MBC is more than 6.4 mg/mL; The FICI of the combined effect of the two substances is 0.5. The 8PL (2 MIC_Ph+2 MIC_La) synergistic combination showed the best spores killing effect in the synergistic combinations, and in the culture medium, the number of spores decreased by 6.1 log CFU/mL after 2 days of treatment in the 8PL synergistic combination; Inoculating 4 log CFU/mL of Alicyclobacillus acidoterrestris spores in apple juice, the 8PL synergistic combination treatment can reduce the spores to 0 CFU/mL after 1 day and maintain the effect for at least 14 days; For apple juice samples with an inoculation concentration of 6 log CFU/mL, the 8PL synergistic combination treatment required more than 3 days to reduce spores to 0 CFU/mL, but after 14 days, 2.2 log CFU/mL spores were detected again. As the concentration of synergistic combination increases, the spore absorbance significantly decreases and the fluorescence spectrum increases; By scanning electron microscopy, irregular wrinkles on the surface of the spores disappeared, and the surface gradually became smooth with an increase in the degree of depression and fragmentation. By transmission electron microscopy, it was observed that the spore membrane cracking; The leakage concentration of nucleic acids and proteins was gradually increasing. The above results indicate that the synergistic combination will cause certain damage to structures such as spore membranes. In addition, the release of pyridine dicarboxylic acid (DPA) also significantly increased, greatly reducing the stress resistance of spores. In summary, the synergistic combination of phloretin and lauraldehyde has a good killing effect on Alicyclobacillus acidoterrestris spores, and it is an effective strategy for fruit juice antibacterial and preservative properties.

SD rats (Sprague Dawley rats) were selected as the research object to investigate the effect of bupropion, an antidepressant, on the antibiotic resistance of gut microbiota. 16S rRNA high-throughput sequencing and metagenomic sequencing technology were used to analyze the effect of bupropion on the gut microbiota and its resistome of rats' fecal and cecal contents and explore the association between them. The results showed that prevalent types of ARGs in the feces and cecum samples included bacitracin, tetracycline, vancomycin, and macrolides-lincomycin-streptogramin (MLS). Bupropion gavage intervention significantly affected rats' total ARG abundance, beta diversity, ARG types, and ARG subtypes of resistome. Gavage administration of bupropion (Bup-PO) increased total ARG abundance in rat feces and cecum contents samples compared to the control (HC) group, but a significant difference was reached only in fecal samples. In terms of ARG types, compared to the HC group, the Bup-PO group significantly increased the relative abundance of 6 ARG types including aminoglycoside, bacitracin, mupirocin, rifamycin, tetracycline, and vancomycin in fecal samples，and also increased the number of vancomycin-resistant genes. In cecum samples, the Bup-PO group significantly increased the relative abundance of 3 resistance gene types, tetracycline, daunorubicin, and fosfomycin. On ARG subtypes, in fecal samples, compared to the HC group, the Bup-PO group significantly increased the relative abundance of vancomycin (vanAG, vanRI, vanSA, vanSI), tetracycline (tetM, tetO, tet32), bacitracin (bceA, bcrA) and rifampicin (rpoB) resistance genes. In cecum samples, the effect of gavage bupropion on ARGs differed from that in fecal samples, with gavage administration of bupropion causing fluctuations in the relative abundance of the MLS class of resistance genes, decreasing the abundance of lmrB but increasing the abundance of macB. Also, gavage administration of bupropion increased the relative abundance of tetracycline (tetW) and daunorubicin (drrA) resistance genes. The above results suggest that bupropion intervention has the risk of increasing antibiotic resistance in rat’s gut microbiota. UCG-005 and norank__f__norank__o_Clostridia_UCG-014 are the major bacterial genera of the rat intestinal flora, and correlation analyses suggest that UCG-005 may be a potential host for tetracycline, rifampicin, mupirocin, and bacitracin, vancomycin resistance genes, while norank_f__norank__o_Clostridia_UCG-014 may be a potential bacterial host for daunorubicin resistance genes, and their increase after bupropion administration may be responsible for the increased abundance of these six types of resistance genes.

Using Gnaphalium affine as the raw material, total flavonoids were extracted via a deep eutectic solvent-assisted heating method. The optimal macroporous adsorption resin was systematically screened, and the purification parameters were subsequently optimized. LC-MS/MS analysis was employed to characterize the chemical composition of purified flavonoids through qualitative and quantitative profiling. An oxidative stress model induced by H₂O₂ in HepG2 cells was employed to investigate the protective effects of the purified Gnaphalium affine total flavonoids against cellular oxidative damage. The results show that, through screening, D101 was identified as the most suitable resin for purifying Gnaphalium affine flavonoids, with a 60% ethanol solution as the elution solvent. The dynamic adsorption-desorption process determined the optimal loading volume and eluent volume to be 124 mL and 200 mL, respectively. Eleven flavonoid compounds were isolated and identified from the total flavonoids of Gnaphalium affine, primarily including luteolin, hyperoside, quercetin, apigenin, and scutellarin, among others. Among these, hyperoside was found to be the most abundant, with a content of 391.91 ± 40.69 μg/g. The purified Gnaphalium affine total flavonoids significantly enhanced cell viability under H₂O₂-induced oxidative stress.  These flavonoids effectively scavenged excess ROS, reduced LDH release, and upregulated SOD and CAT. Additionally, they increased GSH content, demonstrating potent antioxidative and cytoprotective effects. Furthermore, quantitative PCR results indicated that Gnaphalium affine total flavonoids could mitigate oxidative stress by modulating the Keap1/Nrf2 signaling pathway. These findings provide a reference for the development of antioxidant functional products derived from Gnaphalium affine.