Abstract:In order to accurately analyze the morphological characteristics and distribution properties of aggregates in the mixture, and to have a more comprehensive, in-depth, and specific understanding of aggregates, CT scanning, digital image processing, and three-dimensional geometric reconstruction technologies were used to reconstruct the real shape of aggregate particles. Five morphological characteristic parameters of aggregates were proposed, and a digital evaluation and experimental analysis were conducted on the morphological characteristics of three aggregates. The accuracy of the digital reconstruction method was validated and the morphological distribution characteristics of aggregate particles were analyzed. Additionally, the Pearson correlation method was utilized to analyze the correlations among the morphological parameters. The study demonstrates that the use of CT scanning technology and digital reconstruction technology can accurately restore the morphological characteristics of aggregate particles and obtain morphological parameters. There are significant distribution characteristics of the morphology of different particle sizes of the same aggregate. The three-dimensional needle-like index and three-dimensional texture index show little variation across different particle sizes. As the particle size increases, the variability of the morphology parameter values for aggregates decreases. The overall three-dimensional texture index follows a power-law distribution, and the complexity gradually decreases with the increase of particle size. Additionally, as the particle size increases, the three-dimensional edge angle gradually stabilizes. There is a strong negative correlation between three-dimensional edge angle and solid moment degree, as well as between solid sphericity and three-dimensional edge angle. Conversely, there is a strong positive correlation between sphericity of the actual shape and the three-dimensional texture index. Digital 3D reconstruction can accurately and comprehensively describes and analyzes the morphology and distribution characteristics of aggregates.

Abstract:In complex geological environments such as deep layers, the mechanical and damage characteristics of rocks have a decisive impact on the development of high-temperature engineering. To further explore the mechanical properties of high-temperature rocks and their damage mechanisms under load, this study delves into yellow sandstone samples exposed to varying temperatures (25 ℃, 200 ℃, 400 ℃, 600 ℃, 800 ℃). Based on X-ray tomography (CT) technology, obtain internal pore data and 3D model of yellow sandstone, analyze the variation law of porosity of yellow sandstone with temperature. Additionally, numerical simulations were executed to delve into the evolution of microcracks and the damage mechanisms inherent in yellow sandstone under distinct temperature conditions. This microscopic approach unveils the thermal damage mechanisms of rocks under high temperatures. Key findings include: as temperature rises, the total porosity of yellow sandstone follows a quadratic growth trend, accompanied by a decrease in pore distribution uniformity. The main factors of thermal damage in yellow sandstone include: high-temperature dehydration, thermal decomposition of mineral components, and expansion of mineral particles. The increase in porosity due to thermal decomposition and particle expansion is a key factor in thermal damage. Between 25400 ℃, differential expansion and compression of mineral grains generate localized stress zones, predominantly fostering intergranular cracks within yellow sandstone. In the 400-800 ℃ range, phase transitions and mineral component decomposition within yellow sandstone amplify these stress zones, favoring intragranular crack propagation. A damage evolution model of yellow sandstone under thermal action was constructed by defining the damage variable based on the porosity of yellow sandstone, which can provide theoretical basis and technical support for the study of damage mechanism in high-temperature rock mechanics.

Abstract:In order to establish an identification model of rock mass grade reflecting the relationship between TBM tunneling parameters and rock mass categories, improve model building efficiency and recognition rate, a research was carried out on background of a tunnel project. The rock mass characteristics were surveyed and graded based on BQ method and RQD, the TBM working data was collected and the main excavation parameters related to the change of rock mass characteristics were screened. The relationship between TBM tunneling parameters and rock mass grade was fitted based on Light Gradient Boosting Machine (LightGBM) algorithm, and the hyperparameters of LightGBM were optimized using genetic algorithm (GA), then a GA-LightGBM model of rock mass grades identification was established. Results: The accuracy of the GA-LightGBM recognition model reached 93.5%, which was higher than that of the support vector machine model and the random forest model. The model training speed is 8 times faster than the gradient boosting decision tree algorithm. Five TBM tunneling parameters were related to rock strength and rock mass integrity, and the total propulsion force could be used as the main criterion for sensing rock mass characteristics. The study provides an efficient method for analyzing TBM excavation parameters and accurately identifying rock mass grades, providing support for rapid on-site perception of rock mass grades and real-time adjustment of operating parameters.

Abstract:In order to improve the real-time detection and evaluation accuracy of intelligent compaction (IC) quality, a continuous compaction quality prediction method based on GA-XGBoost model was proposed to improve the prediction accuracy of dynamic deformation modulus (E_vd). The model takes the dynamic deformation modulus as the goal, establishes a machine learning model, mainly uses the decision tree algorithm, and constructs the XGBoost model to predict and analyze the compaction quality. In order to improve the prediction accuracy and reliability of the model, genetic algorithm (GA) is introduced to optimize the hyperparameters of the model. Firstly, through the field engineering test, the vibration acceleration of the roller is measured, the acceleration signal is analyzed, the signal statistics are calculated and the harmonic frequency is obtained by fast Fourier transform (FFT), and the system connection between the characteristic factors and E_vd is preliminarily established. Secondly, the characteristics of each time-frequency domain are screened, the correlation analysis is carried out, and the characteristics with high correlation are selected to establish the prediction model. Finally, it is verified that the GA-XGBoost prediction model can better predict E_vd.The results show that the genetic algorithm (GA) can efficiently determine the hyperparameters of the XGBoost algorithm, and it shows better convergence speed than the single XGBoost model. By optimizing the feature factors and changing the input parameters, the prediction accuracy of the GA-XGBoost model is improved. The optimized mean square error is 3.9% and the correlation coefficient is 0.748. At the same time, compared with the traditional CMV fitting E_vd method, the machine learning model can greatly improve the prediction accuracy.

Abstract:Accurate vehicle dynamic parameter identification is a prerequisite for vehicle-bridge coupling vibration analysis and bridge health monitoring. This study proposes a rapid identification method for the vehicle dynamic parameters based on complex modal analysis and multi-core parallel genetic algorithm. Firstly, an algorithm combining the complex modal theory with the finite element method is improved to calculate the natural frequencies, damping ratios, and modal shapes of vehicles. Subsequently, a multi-core parallel genetic algorithm for the vehicle dynamic parameter identification is proposed, in which the fitness evaluation of multiple objectives including natural frequencies, damping ratios and mode shapes is adopted, and the dynamic parameters including stiffness coefficient, damping coefficients and moments of inertia of the vehicle model can be rapidly and accurately identified. Finally, the wheel-drop-vibration experiment and ambient-excitation peak method are used in modal analysis of practical vehicle to obtain the measured natural frequencies, damping ratios and vibration modes, which are used in the fitness evaluation. The above methods have been validated by the dynamic parameters identification of the practical light car and heavy truck, and the results show that: vehicle vibration modes calculated with the identified vehicle dynamics parameters are in good agreement with the measured vibration modes, in which the maximum error of the first three natural frequency is 0.8%, the maximum error of the damping ratio is 1.3%, and the cosines of the angle between the calculated and measured vibration mode vectors are close to 1; incorporating body torsional damping is critical to accurately capture the torsional vibration characteristics of real vehicles; the multi-core parallel algorithm greatly accelerates the identification process. The acceleration ratio of 16-core CPU reaches the maximum value of 12.5 when 15 cores are in parallel. Therefore, the multi-core parallel algorithm is very effective in multi-objective and multi-parameter identification of complex vehicles.

Abstract:To evaluate the impact of predictive-forward-collision-warning (PFCW) on the driving safety of the human-driven leading vehicle during rapid deceleration of the preceding vehicles on freeways, this study designed a connected mixed platoon featuring a human-driven leading vehicle and developed a connected human-machine interface (HMI) with PFCW functionality. Thirty-six participants were recruited for driving simulation experiments that considered platoon configuration (single vehicle or platoon) and connectivity condition (with or without). By selecting driver behavior characteristic indicators, surrogate safety measures, and vehicle operation indicators, the impact of PFCW was analyzed in terms of drivers′ longitudinal risk avoidance behavior, dynamic collision risk, and overall safety of the mixed platoon. The results indicated that PFCW enhances drivers′ longitudinal risk avoidance capabilities and reduces the collision risk of the leading vehicle. The leading vehicle in the connected mixed platoon mode demonstrated the best performance in terms of driver safety, ultimately improving the overall safety of the mixed platoon. The findings provide valuable insights for optimizing PFCW information, developing navigate on autopilot system, and promoting the implementation of connected mixed platoons.

Abstract:It is essential to accurately capture traveler preferences for developing sustainable operating strategies with demand-responsive transit (DRT) services. Considering the unmatured practice of DRT and limited public awareness in China, the loss aversion psychology exhibited by travelers when faced with complex and uncertain travel scenarios is focused. Integrating socioeconomic attributes, travel characteristics, and modal attributes, a mode choice model based on the hybrid utility and regret model is developed and calibrated through a dataset comprising 15 012 observations from Beijing. The results indicate travelers aged 60 and above show the most pronounced preference towards adopting the DRT. Factors such as female, age 25 and above, educational attainment at the associate degree level or higher, monthly disposable income exceeding 1 000 yuan, and household sizes of two or more individuals have a positive influence on the adoption of the DRT. Furthermore, under the travel distance within 10 kilometers, the fare for DRT is suggested to not exceed 10 yuan, and the vehicular detour time is suggested to be limited within one-fourth of the total travel time to fully utilize DRT′s potential in augmenting the modal share of sustainable transportation.

Abstract:The stability of embankments in regions with heterogeneous soft soils presents a key technical challenge in geotechnical engineering. To enhance the assessment of their three-dimensional (3D) stability, a 3D basal failure mechanism for reinforced embankments on heterogeneous soft soils is developed based on the upper bound limit analysis theorem, and the corresponding energy conservation equation is formulated using the principle of virtual work. A genetic algorithm is introduced to construct an efficient optimization strategy for solving the upper bound solution of 3D stability. The proposed failure mechanism is further degenerated into a 3D toe failure mode of slopes and compared with existing upper bound solutions for slopes, thereby verifying the accuracy and computational efficiency of the genetic algorithm. On this basis, a series of parametric analyses are conducted to investigate the influence of stratification characteristics, unsaturated strength properties, heterogeneous shear strength distribution, tensile strength of reinforcement, number of reinforcement layers, and matric suction on the 3D stability of reinforced embankments. The results indicate that under the 3D basal failure mechanism, a larger ratio of heterogeneity coefficients between the unsaturated subsoil and the embankment significantly improves overall stability. For a fixed reinforcement tensile strength, reducing the reinforcement spacing weakens the stabilizing effect of matric suction. When the number of reinforcement layers remains constant, the stabilizing contribution of reinforcement becomes more pronounced with increasing soil heterogeneity. The proposed 3D stability analysis method offers a reliable computational tool for optimizing reinforcement configurations and supports rational embankment design under complex foundation conditions.

Abstract:While deep learning has advanced traffic object detection, accurately detecting small objects in complex traffic scenes with dense occlusion remains challenging. To address these issues, this paper proposes a novel small traffic object detection algorithm, CDAQ-DDETR, which incorporates an attention-based deformation and dynamic querying mechanism. Building upon Deformable DETR, the algorithm introduces the CBAM attention-based dual-tower mechanism and DCNv2 Deformable convolutions to reconstruct the original residual network, thereby enhancing the semantic acquisition capabilities for small traffic objects in dense areas. By leveraging the AFN network concept to add lower-level features and constructing an attention-aware fusion pyramid module, the algorithm improves detection performance for multi-scale small and medium traffic objects. Additionally, by integrating a dynamic query mechanism module before the original decoder, combined with matching input image characteristics, it constructs optimal query vectors, enhancing the algorithm′s adaptability and generalization ability against diverse background interferences. Experiments conducted on the VisDrone2019 dataset show that the CDAQ-DDETR algorithm has achieved a mean Average Precision (mAP@0.5:0.95) of 37.9% and a mean Average Recall (mAR@0.5:0.95) of 57.4%. Compared to the current state-of-the-art (SOTA) algorithms, there is an improvement of 5.5% in detection precision and 8.0% in recall rate, particularly, an increase of 6.9% in precision and 10.0% in recall rate for detecting small objects. Visualization experiments further demonstrate its practical applicability and superior performance in detecting small traffic objects in dense scenes.

Abstract:In response to the problem of difficult control of shield tunneling attitude in subway tunnels, taking a tunnel project in Changchun as an example, a shield tunneling attitude prediction model (EMD-LightGBM) was constructed based on on-site measured data, which integrates empirical mode decomposition (EMD) and lightweight gradient boosting machine (LightGBM). Firstly, filter the features of the original dataset through feature importance and correlation analysis. Then, the data is decomposed into multiple stationary subsequences and combined into a new dataset by EMD. Finally, EMD-LightGBM was fitted and trained by the new dataset to achieve the prediction of shield tunnel attitude, and the prediction performance of the model was compared with that of LightGBM alone and the EMD-BPNN. Verify the excellent performance of the EMD-LightGBM model through two evaluation systems: prediction accuracy and prediction stability. The results showed that compared with LightGBM and EMD-BPNN, EMD-LightGBM performed the best in predicting shield attitude deviation in the line graph, with a maximum mean absolute error (E_MA) and root mean square error (E_RMS) of 2.89 mm and 4.13 mm, respectively, and a minimum coefficient of determination R² of 0.95. Meanwhile, the maximum 95% confidence intervals for the E_MA and mean square error (E_MS) of EMD-LightGBM predictions are 3.5 mm and 25.6 mm², respectively. Combined with the good frequency distribution of its predicted absolute error (E_A) and square error (E_S), it demonstrates the high accuracy and stability of EMD-LightGBM in predicting shield tunnel attitudes. The research results can provide a theoretical method for the attitude control of shield tunneling in similar projects.

Abstract:To optimize the vibration control capability of concrete, this paper proposes several methods to enhance the damping performance of steel fibers through surface modification, thereby improving the vibration reduction effect. Steel fibers were surface-modified using silane coupling agent, emulsified asphalt, and polyurethane. The effects of steel fiber surface treatment on the time-domain curve, damping ratio, energy dissipation factor, and microstructure of concrete were investigated via vibration excitation method and scanning electron microscopy (SEM), and the influencing laws and mechanisms were analyzed.The results show that all three surface treatments can increase the damping ratio of steel fiber-reinforced concrete, with the maximum damping ratio achieved when the steel fiber volume content is 1%. The emulsified asphalt-treated group exhibits the most significant improvement in vibration reduction performance: at a 1% volume content, it maintains a compressive strength of 50.5 MPa while increasing the damping ratio to 221% of the control group. The emulsified asphalt layer on the steel fiber surface enhances the vibration energy dissipation of the matrix, thereby improving the damping performance of concrete.The interface modulation method between steel fibers and the matrix determines the vibration reduction mechanism: silane coupling agent treatment enhances the bonding force between fibers and the matrix, improving energy dissipation during vibration by increasing slip energy consumption at the fiber-matrix interface; polyurethane treatment enhances the deformability and viscoelasticity of the fiber-matrix interface, thereby increasing energy dissipation during vibration propagation; emulsified asphalt treatment achieves energy dissipation through both mechanisms simultaneously.This study provides a concrete mix ratio that meets the workability and strength requirements in practical construction while exhibiting excellent vibration reduction performance. As a new vibration reduction method for subway ballast bed materials, it can serve as a reference for future engineering practices and research.

Abstract:In order to accurately simulate the interaction between concrete deck slabs and steel girders of steel-concrete cable-stayed bridges and the redistribution of internal forces between the two due to the shrinkage and creep of concrete, according to the internal force distribution characteristics of the composite cross-section and the concrete shrinkage and creep recursive algorithm, a new type of continuous composite girder virtual double-layer beam element was developed. Then the new element was validated and analyzed by using three-span continuous girders and the Chibi Yangtze River Highway Bridge project, with a comparative study conducted against the actual double-layer element commonly used in engineering. The results show that the calculation results of the virtual double-layer beam model are in good agreement with the measured data, and the number of elements can be reduced by about 67% and the calculation speed can be improved by about 40% compared with the actual double-layer model. The new element exhibits high computational accuracy while significantly reducing main girder elements count. It realizes the continuous composition between the interface of steel girder and bridge deck, accurately simulates the shrinkage and creep effects of concrete bridge deck, and also simulates the construction process of steel girder and bridge deck in stages, which provides an improved steel-concrete composite beam simulation method of high accuracy for the composite structure.

Abstract:In order to explore the characteristics of microstructural changes of granite residual soil during wetting, the electron microscope scanning test was used to analyze the pore changes of the soil under unsaturated-saturated state by PCAS software to achieve quantitative characterization of the pore structure, and the three-dimensional fractal dimension characteristics were further analyzed by analyzing the T₂ of the soil under unsaturated-saturated state by nuclear magnetic resonance test. The research results show that under unsaturated state, with the increase of water content, the number of smaller pores decreases and the number of larger pores increases. In this process, small aggregates absorb coarse particles to expand and form larger aggregates; after reaching the saturated state, the aggregates disintegrate and show the opposite development trend; reaching the saturated state. Based on the transverse relaxation time (T₂), the pores can be divided into three categories: aggregate pores, intergranular pores, and microcracks. Among them, intergranular pores are most significantly affected by pore water, while aggregate pores and microcracks are less affected by pore water; the three-dimensional structural fractal dimension shows a decreasing trend with the increase of water content, and has an exponential function relationship with the water content. The study on the change law of microstructure during the wetting process of granite residual soil provides a theoretical basis for analyzing the initiation mechanism of mass landslides of granite residual soil. The microstructure evolution stage can be predicted by monitoring the changes in soil moisture content. The slope stability evaluation index can be established by combining the fractal dimension model, which provides a key criterion for early warning of granite residual soil landslides in the hilly areas of Fujian and Guangdong.

Abstract:In order to improve the ductility seismic performance of railway piers in high intensity seismic regions, then reducing the internal force at the bottom of piers under strong earthquakes and protecting the pile foundations, this paper presents two new ductile structure schemes for railway pier. And low cyclic loading tests were carried out on a prototype pier model specimen and two new pier model specimens with ductile structure. Finally, the damage behavior, failure mechanism, hysteretic performance, energy dissipation capacity, residual displacement, stiffness degradation characteristics and strength degradation characteristics of different piers were analyzed. The test results show that the failure mode of round-ended hollow pier commonly used in railway is mainly bending failure. In the plastic hinge area at the bottom of the pier, a large number of cracks develop, concrete flake off, stirrups are exposed, and some longitudinal reinforcement fracture. But the failure modes of two new piers with ductile structure are closer to the corner failure of the rocking pier. Their failure area is mainly concentrated in the range of 20 cm at the bottom of the piers. By the end of the tests, only six to nine horizontal cracks appear on the piers, and partial concrete at the edge of the piers bottom is crushed and peeled off. The initial stiffness of the new piers with ductile structure can be consistent with that of the prototype pier to meet the stiffness requirements of railway piers and ensure the operation safety of trains. The equivalent yield forces of the new piers 2 and 3 are 91.31% and 77.12% of that of the prototype pier, and the ultimate horizontal forces at the top of the piers are 72% and 65% of that of the prototype pier. The internal forces of pier bottom and pile foundation in the new piers with ductile structure can be effectively reduced under strong earthquakes, thus protecting the pile foundation. The equivalent viscous damping ratios of the new piers are more than 50% of that of the prototype pier, and the ductility seismic performance of railway round-ended hollow pier is improved through new ductile structure schemes.

Abstract:To effectively address the impact of catastrophic weather on urban rail transit passenger flow organization and to explore the rules of passenger flow changes under such weather conditions, this paper conducts a prediction study on urban rail transit network passenger flow. Based on a deep residual network (DRN) and bidirectional long short-term memory (BiLSTM), a DRN-BiLSTM prediction model incorporating catastrophic weather features is constructed. The model′s performance is evaluated using metrics such as mean squared error (MSE), root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R²), and its passenger flow prediction effectiveness is verified and analyzed. The results show that compared to traditional LSTM and BiLSTM, when inputting catastrophic weather features, the DRN-BiLSTM model reduces MSE by 22.10% and 21.96%, RMSE by 10.54% and 10.46%, MAE by 3.20% and 3.95%, and increases R² by 5.01% and 2.12%, respectively. By optimizing model parameters with the grid search method, the model training loss is reduced by 36%. Practical verification demonstrates that the DRN-BiLSTM combined model constructed in this paper can effectively capture deep data features and significantly improve the accuracy of passenger flow prediction.

Abstract:In order to comprehensively improve the effect of traffic accident severity prediction, for the current stage of traditional machine learning and deep learning methods with limited prediction accuracy and slow convergence of the network, proposing an improved FBLS method for predicting accident severity at urban road intersection. The model replaces the feature node layer of BLS with Takagi-Sugeno fuzzy system to extract the hidden features of high-dimensional accident data more extensively and still retains the fast convergence characteristics of BLS; the SMOTE algorithm is also fused in the input layer of the FBLS to balance the accident data categories and enhance the reliability of the prediction results. Through the historical data of traffic accidents in Greater Manchester, UK, the original FBLS was selected in the horizontal dimension, and RF, SVM, BPNN, LSTM, CNN, which are commonly used for traffic accident severity prediction, were chosen in the vertical dimension, to compare the model performance with the S-FBLS. The results show that comparison with the original FBLS, S-FBLS improves the accuracy of severe accidents by 52.87%, comparison with five comparative models, S-FBLS improves the network training speed by more than 97%, improves the overall accuracy by 2.2%, 8.95%, 8.68%, 6.47%, 5.64%, improves the specificity by an average of 6.49%, improves the sensitivity by an average of 6.31%, and improves the precision by an average of 5.66%. The S-FBLS-driven accident severity prediction method can provide a reliable theoretical support for the early warning of the occurrence of accident at urban road intersection.