交通运输工程学报

Cover and Contents of Vol.24, No.6, 2024

Cover Contents

Review on key emergency resource deployment for expressway traffic accidents

JIN Yin-li, GAO Peng, TAN Er-long, LI Lin-wei, LIU Kun

Abstract: More> The key emergency resources for expressways were defined. In terms of the prevention, preparation, and response stages of emergency rescue for expressway traffic accidents, the key emergency resource deployment for expressways was divided into emergency facility location selection, allocation and dispatching of key emergency resources. The research achievements in the three aspects were systematically reviewed, and the existing problems and future research directions were discussed. Research results show that in terms of the location selection for emergency facility, current research is mostly oriented to the initial scenario of expressway planning, and the location selection result is relatively fixed. It is necessary to study the location selection methods for small and miniature emergency facilities oriented towards the stage of road operation. Limited by the model’s solving performance, existing research seldom considers the location selection for large-scale emergency facility. The combined solving algorithm that integrates multiple optimization algorithms is expected to make breakthroughs. In terms of the key emergency resource allocation, current research assumes that the complete accident information is available during the initial phase of the accident, which is inconsistent with actual condition. Using a static allocation scheme to design a dynamic resource allocation strategy is more realistic. In terms of the key emergency resource dispatching, uncertainty after traffic accidents brings challenges. Robust methods estimating the traffic state should be studied. The time-varying characteristics of traffic state after traffic accidents have great impact on the speeds of emergency vehicles. The key emergency resource dispatching methods integrating the dynamic path planning and traffic control strategies should be studied. Looking toward the future, it is necessary to study the key emergency resource dispatching method in the environment with mixed traffic flow and intelligent and connected vehicles. The dynamic change law of vehicle speed in the process of emergency vehicle dispatching should be deeply explored. In terms of the integration, the research on the integrated optimization of expressway scenarios has not yet been studied. Therefore, it is necessary to study the theory and verification methods of key emergency resource deployment for land-sea-air-space oriented towards special scenarios, such as bridges and mountainous areas, so as to improve the resilience of expressway transportation network.

2024, 24(6): 1-25. doi: 10.19818/j.cnki.1671-1637.2024.06.001

FullText HTML (31) PDF (10518KB) (13)

Review on mechanism and key technologies for delaying thermal crack propagation of high-speed train brake discs

LI Jie, WANG Li, WANG Xiao-yan, HU Zheng, WANG Xin, GAO Zi-yu, WANG Shi-min

Abstract: More> The fatigue thermal crack initiation and propagation mechanisms of train brake discs made of cast steel were analyzed. The effects of stress ratio, temperature, and overloading on the crack propagation rate of brake discs were described. Three key technologies for delaying crack propagation, including shot peening, laser cladding, and cold spraying, were systematically introduced. Future research directions for mitigating surface crack propagation in brake discs were proposed. Analysis results show that frequent braking and releasing in trains subject the disc surface to alternating thermal stress so that the formation of microcracks was induced at grain boundaries and defect sites within material. These microcracks subsequently propagate into radial cracks and eventually evolve into circumferential cracks. Microcracks are found to propagate radially under localized frictional forces, merging with primary radial cracks and increasing their length. It ultimately leads to brake disc failure. Plastic deformation of the disc surface is induced by shot peening and cold spraying treatments, transforming the residual tensile stress into residual compressive stress. Therefore, the propagation of fatigue cracks is slowed down. The microstructural properties of the base metal of brake disc are improved through laser cladding, enhancing its wear resistance at elevated temperatures. In the future, in-depth research should be carried out on the influence of wear on the propagation of surface cracks in brake discs, the mechanisms of grain refinement and dislocation density evolution during shot peening, and the correlation and quantitative relationships among shot peening parameters, such as shot diameter, coverage rate, and peening intensity. The application efficiency of high-speed laser cladding in the rapid prototyping of friction pairs should be further explored. Additionally, for cold spraying technology, attention should be paid to how to enhance the driving force for plastic deformation by increasing particle velocity and temperature, thereby reducing microstructural defects in the deposition layer.

2024, 24(6): 26-42. doi: 10.19818/j.cnki.1671-1637.2024.06.002

FullText HTML (21) PDF (12292KB) (12)

Review on travel quality evaluation methods for urban bicycle traffic system

LI Cong-ying, ZHANG Hong-tao, LI Kun, ZHANG Da-peng, JIA Jin-xiu, ZHAO Song-yang, HE Yuan

Abstract: More> In order to investigate the influencing factors of travel quality for urban bicycle traffic system, a comprehensive review of the evaluation methods was carried out from the aspects of both road facilities and road networks. In the evaluation of road facilities, factors such as bicycle traffic flow, bicycle lane design elements, bicycle travel environment, and cyclist perception were considered, and evaluation methods such as bicycle lane passing capacity, bicycle lane service level, bicycle lane safety evaluation, cyclist stress, and cyclist satisfaction were established. In the evaluation of road networks, topological analysis methods such as complex networks and space syntax were applied, and evaluation methods such as bicycle accessibility and bikeability were established. Research results show that in the evaluation method of road facilities, the evaluation subject shifts from bicycles to multimodal transportation, and the evaluation metrics consider the impact of motor vehicles, buses, pedestrians, and other factors on cycling. The evaluation perspective transitions from focusing on road designers to cyclists, progressively substituting designer experiences with cyclist perceptions to determine evaluation levels. The data of cyclist perception is obtained by questionnaires, laboratory videos, field experiments, and virtual environment experiments, and the modeling methods are mainly discrete choice models, statistical analysis, linear regression models, and structural equation models. The research focuses on the measurement method and influencing mechanism of psychological perception. The influencing mechanism of the physiological perception of cyclists still needs to be studied in depth, and the influencing mechanism needs to be refined by taking into account individual differences. In the evaluation method of road network, the topological analysis method mainly based on complex network and spatial syntax verifies the influence of the topological relationship of road networks on the travel frequency of cyclists. Bicycle accessibility takes into account the influence of cycling distance and travel destinations on cycling, and bikeability comprehensively considers the influence of road facilities and road network structure on cycling demand. The synergistic mechanism of road facilities and road network characteristics still needs to be studied in depth. In the future, it is necessary to improve the evaluation system for road facilities of bicycles at all stages and establish synergistic evaluation and optimization methods that consider both road facilities and road network structure, and provide theoretical references for improving the travel quality of bicycle traffic systems.

2024, 24(6): 43-65. doi: 10.19818/j.cnki.1671-1637.2024.06.003

FullText HTML (18) PDF (3016KB) (10)

Research review on behavior strategies of electric vehicles considering charging demands

MA Jun-chi, ZHANG Yuan, DUAN Zong-tao, TANG Lei

Abstract: More> In order to improve the usability and operational efficiency of electric vehicles, the research progress of related work was described from three perspectives: charging behavior strategies (including charging station recommendation and charging path planning), behavior strategies for passenger service (including ride-sharing and car rental scenarios), and behavior strategies under vehicle-to-grid (V2G) interactions. The principles and applications of artificial intelligence technology were summarized, and the future research directions were explored. Research results show that the research on charging station recommendation focuses on two optimization objectives: time overhead and charging fee. Heuristic algorithm or reinforcement learning algorithm is often applied to obtain the optimal charging station. Charging path planning needs to construct path energy constraints and energy recovery mechanisms according to the characteristics of electric vehicles. In general, the Pareto optimal method or reinforcement learning algorithm is used to optimize the path with time, energy, and other objectives. In ride-sharing scenarios, the research on behavior strategies mainly uses the temporal and spatial distribution features, and coordinate order dispatch, charging, and repositioning operation to maximize fleet profit. In car rental scenarios, the research on behavior strategies uses charging and repositioning operation to provide abundant available electric vehicles to satisfy users' needs at service stations. Research on behavior strategy in V2G scenarios focuses on the three optimization objectives of charging/discharging cost effectiveness, power grid stability, and energy utilization efficiency. Mathematical programming method or reinforcement learning algorithm is often used to optimize the charging/discharging behavior of electric vehicles. Future research on behavior strategies of electric vehicles should focus on the changes in charging behavior after the introduction of autonomous driving technology, with attention to the interpretability and scalability of the model. From the system perspective, battery degradation and integrated scheduling should be further considered.

2024, 24(6): 66-79. doi: 10.19818/j.cnki.1671-1637.2024.06.004

FullText HTML (22) PDF (1164KB) (9)

Seismic performance of large-diameter and variable cross-section pile group foundation in earthquake-induced subsidence sites

FENG Zhong-ju, XU Bo-xi, DONG Jian-song, ZHANG Cong, LIU Xu-zhao, LAI De-jin

Abstract: More> The model test on shaking table was used to select synthetic 5010 waves with different ground motion intensities. The effect of different ground motion intensities on the seismic performance of pile group foundation was studied. The differences in dynamic responses of horizontal displacement of pile top, bending moment of pile body, and pile foundation damage between single pile and pile group foundations under different ground motion intensities were compared. Research results indicate that under different ground motion intensities, pile group foundation is more stable than single pile foundation and less disturbed by seismic waves. The peak horizontal displacement of pile top of pile group foundation is significantly smaller than that of single pile, and the difference increases with the increase in ground motion intensity, with a maximum of 1.15 mm. Moreover, the peak horizontal displacement of pile top of pile group foundation occurs later than that of single pile. The bending moment change of pile group foundation along the pile body is the same as that of single pile, which increases first and then decreases. The difference in mechanical properties between silty soil layer and non-silty soil layer is more obvious due to earthquake-induced subsidence, so the peak value reaches at the interface of silty soil layer. The peak bending moments of pile body of pile group foundation under different ground motion intensities are smaller than those of single pile foundation, and the maximum difference is 13.02 kN·m. The peak bending moment of pile body of pile group foundation occurs later than that of single pile foundation. It can be concluded that there is a significant difference in the dynamic response between pile group foundation and single pile foundation under different ground motion intensities. The ground motion intensity that pile group foundation can withstand when damage occurs 0.35g, while that for single pile is 0.30g. The fundamental frequency drop of pile group foundation is 27.23%, while that of single pile is 33.46%. It indicates that earthquake-induced subsidence site has less influence on the fundamental frequency of pile group foundation, and pile group foundation can withstand greater ground motion intensity without damage. In summary, the seismic performance of pile group foundation is better than that of single pile in earthquake-induced subsidence sites. In engineering practice, the seismic performance of pile foundation can be improved by considering pile group effect and reasonable design of pile type.

2024, 24(6): 80-91. doi: 10.19818/j.cnki.1671-1637.2024.06.005

FullText HTML (21) PDF (12844KB) (8)

Calculation method of instantaneous stiffness of steel reinforced ultra-high performance concrete beams with rectangular section

LIN Shang-shun, JI Bang-chong, LIU Jun-ping, LIN Jian-fan, ZHAO Jin-bing

Abstract: More> To study the instantaneous stiffness and calculation method of steel reinforced ultra-high performance concrete (SRUHPC) beams with rectangular section, five SRUHPC beam specimens and one reinforced ultra-high performance concrete (reinforced UHPC) beam specimen were fabricated. The reinforcement rates of the specimens ranged from 0.8% to 1.1%. The built-in structural steel was in the shapes of I-beam, inverted T-beam, and H-beam, with steel content ranging from 8.7% to 15.6%. The flexural test was carried out. The deformation patterns of SRUHPC beam specimens with rectangular section and reinforced UHPC beam specimen, and the effect of the change of design parameters on the stiffness of SRUHPC beam specimens, were analyzed. Based on the test result and the stiffness analysis method, the simplified calculation method for the instantaneous stiffness of SRUHPC beams with rectangular section was proposed. Under service loads, the maximum deflections of 13 SRUHPC beam specimens, including 8 specimens from literature, were calculated, and the calculated results were compared with the test values. Analysis results show that, under the same load, compared with the reinforced UHPC beam specimen, the deflections of SRUHPC beam specimens with rectangular section reduce by 16%-72%. The stiffnesses of SRUHPC beam specimens with rectangular section decrease by up to 7% due to section cracking, which is not obvious. Similar to the longitudinal tensile reinforcement, the tensile side portion of structural steel can also resist some tensile force when the SRUHPC beam specimens with rectangular section are loaded, thus inhibiting crack development in the beams and reducing the loss of stiffness due to cross-section cracking. The proposed method for the instantaneous stiffness yields an average ratio of 1.041 for deflection calculation values to test values, and the ratio variance is 0.017. Therefore, these calculated values are in good agreement with the test values, and the proposed method has higher accuracy than the existing calculation methods in literature.

2024, 24(6): 92-105. doi: 10.19818/j.cnki.1671-1637.2024.06.006

FullText HTML (25) PDF (12537KB) (8)

Stress concentration characteristics of concrete-filled steel tubular truss-rib K-joint with inner studs

LIU Jun-ping, YANG Qian, LIU Hua-long, YANG Yi-tu, CHEN Bao-chun

Abstract: More> To further study the influence of setting inner studs on the stress distribution of concrete-filled steel tubular truss-rib K-joints, a test on the stress concentration characteristics was conducted. The hot spot stress and stress concentration factor of joints without inner studs were compared and tested. The fine finite element models of the joints with inner studs were established. The influences of the arrangement and geometric dimensions of inner studs on the stress concentration factor were analyzed. Based on the results of the test and the finite element parameter analysis, the recommended layout of the inner stud and the calculation method of the stress concentration factor of the joints were proposed. Research results show that the hot spot stress distribution of the joints is not changed with the inner studs, and the maximum stress concentration factor is found at the crown point of the main tube in the tensile weld area, but the stress concentration degree of the joints effectively reduces. Compared with the joint without inner studs, the maximum hot spot stress of the joint with inner studs reduces by 17.38%. The overall stress concentration factors of the main tube side and the corresponding branch tube side in the tensile weld area decrease by 24.20% and 12.30% on average, respectively, and the circumferential section stress of the main tube decreases by 16.2%. The influence of inner studs on the stress distribution and stress concentration in the compressive weld area is so slight to be within 7%. The stress concentration factors are affected greatly by the circumferential arrangement angle and the axial arrangement spacing of inner studs but less by the geometrical dimension of inner studs. The inner studs are suggested to be arranged within the range of approximately 2.7 times the diameter of the branch tube from the center of the joint to each side section in the axial direction of the main tube. The circumferential arrangement is within the range of [-60°, 60°] of the central angle of connecting branch tube side. At the same time, the interval between adjacent studs should not be less than 6 times the diameter of the inner stud but not more than 400 mm. The calculation formula of the stress concentration factor introduced by the influence coefficient of axial arrangement and circumferential arrangement of inner studs can be used in evaluating the fatigue performance of concrete-filled steel tubular truss-rib K-joints with inner studs with high calculation accuracy.

2024, 24(6): 106-120. doi: 10.19818/j.cnki.1671-1637.2024.06.007

FullText HTML (20) PDF (11389KB) (9)

Imaging method for damage inentification of plate structures detected by single mode Lamb waves

TIAN Liang, WANG Yu-ning, FAN Li-long, ZHAO Jian, SI Zhi-yuan

Abstract: More> The damage detection of steel plates with prefabricated rectangular through-cracks was carried out by using single mode Lamb waves. The excitation and receiving transducers were arranged in a sparse array on the damaged steel plates. A numerical model of Lamb waves propagation in the steel plate was established based on ABAQUS software, and a corresponding experimental platform was built. Through numerical simulation and experiments, the propagation characteristics of Lamb waves in the steel plate were analyzed. A steel plate damage localization program was developed based on MATLAB. The elliptic trajectory was determined based on differential signals of the excitation and receiving transducer. Then, damage position of the steel plate was determined according to the intersection points of multiple sets of elliptic trajectories. The multiple sets of elliptic trajectories were fused to locate and visualize the steel plate damage. Based on the numerical simulation results and experimental data of the Lamb wave damage detection, the influences of different data fusion methods on the accuracy of steel plate damage localization were further compared. Research results indicate that using an excitation signal frequency of 200 kHz and a bilateral symmetrical excitation method can effectively ensure the single mode characteristics of Lamb waves, thereby avoiding dispersion effects and multimodal interference when Lamb waves propagate. For steel plates with single damage, damage imaging errors for both the amplitude summation method and the amplitude multiplication method are within 5 mm. For steel plates with double damage, imaging results of the amplitude summation method will exhibit an indistinguishable pseudoscopic image, with a damage imaging error up to 30 mm. However, imaging results of the amplitude multiplication method are still close to the true damage location, with a maximum error of only 4 mm. Moreover, the contrast of damage imaging is more obvious, indicating that the amplitude multiplication method has more advantages in predicting multiple damage in the steel plate. The localization error of steel plates with double damages obtained from numerical simulation and experimental signals is within 2 mm, which verifies the good damage prediction accuracy of the numerical model.

2024, 24(6): 121-134. doi: 10.19818/j.cnki.1671-1637.2024.06.008

FullText HTML (14) PDF (16767KB) (6)

Concrete bridge crack detection method based on improved YOLO v8s in complex backgrounds

JIANG Shi-xin, ZOU Xiao-xue, YANG Jian-xi, LI Hao, HUANG Xue-mei, LI Ren, ZHANG Ting-ping, LIU Xin-long, WANG Di

Abstract: More> To address the issue of low detection accuracy for cracks in concrete bridges caused by complex backgrounds, as well as small and obscure features, the crack location information was accurately located, and key parameters such as crack length and width were measured based on the improved YOLO v8s algorithm for concrete bridge crack detection. Based on the YOLO v8s model, the omni-dimensional dynamic convolution (ODConv) was incorporated to capture richer contextual information in feature maps, enhancing the model's ability to extract the target features and detect small and obscure cracks. An improved channel attention module was used to develop the concatenated two-layer feature-modified attention (C2f-MA) fusion module, enabling the extraction of more texture information from feature maps. This modification further made the network focus on crack features, suppressing interferences from irrelevant background information and improving the crack detection performance in complex backgrounds. The weighted intersection over union (WIoU) loss function was introduced to address the challenge of low-quality sample recognition, optimizing the model's convergence speed and detection accuracy. Crack images with complex backgrounds such as small and obscure cracks, shadows, artificial lines, and weeds were screened in the bridge detection report. A bridge crack image dataset was established by manual annotation with complex background conditions. The model's performance was comprehensively evaluated through comparative and ablation experiments by taking recall, average precision and model storage capacity as quantitative evaluation indicators. Research results demonstrate that the improved YOLO v8s algorithm achieves recall, average precision of 0.829, 0.893 and 0.631, respectively, as well as the model storage capacity of 11.14 MB. Its comprehensive evaluation indicators outperforms the baseline YOLO v8s and other target detection models, validating that the proposed algorithm exhibits robust performance in complex backgrounds.

2024, 24(6): 135-147. doi: 10.19818/j.cnki.1671-1637.2024.06.009

FullText HTML (14) PDF (20360KB) (7)

Fault-tolerant control of four-wheel hub-drive electric vehicles based on MPC-MSCA

XUE Hong-tao, ZHANG Yu-le, SONG Zi-wei, FENG Han

Abstract: More> In response to the safety hazards posed by local in-wheel motor failures in four-wheel hub-drive electric vehicles, a fault-tolerant control method based on model predictive control-multi system collaborative allocation (MPC-MSCA) was proposed. The method was designed to address operating conditions characterized by insufficient output capacity following local in-wheel motor failures. A 14-degree-of-freedom (DOF) four-wheel hub-drive vehicle dynamics model was constructed, comprising a 6-DOF vehicle body model and four 2-DOF wheel models. A fault-tolerant control method was designed by referencing a 2-DOF vehicle model, which included a motion tracking layer and a torque distribution layer. The motion tracking layer was utilized to calculate the total longitudinal force and additional yaw moment required for normal vehicle operations. In the torque distribution layer, an optimization allocation scheme and a MSCA scheme were designed to respectively address operating conditions of sufficient and insufficient output capacity following local in-wheel motor failures. The objective function, constraints, and control variables of the MSCA control strategy were highlighted. The proposed MPC-MSCA control method was validated through co-simulation with Simulink/MATLAB and CarSim under two operating conditions: straight-line driving and double lane change driving. Research results show that the method reduces the average yaw velocity error of vehicles by 31.6% compared with the traditional method in a straight-line driving environment, effectively ensuring the vehicle's ability to drive straight when there is a local in-wheel motor failure. In a double lane change driving environment, the average sideslip angle error and average yaw velocity error reduce by 7.4% and 6.9%, respectively, improving the vehicle's handling stability when there is a local in-wheel motor failure. Therefore, the proposed fault-tolerant controller can ensure the four-wheel hub-drive electric vehicle meeting the handling stability and safety requirements under the condition of one or two in-wheel motor failure.

2024, 24(6): 148-158. doi: 10.19818/j.cnki.1671-1637.2024.06.010

FullText HTML (11) PDF (4893KB) (6)

Influence of frequency-dependent characteristics of double-layer nonlinear fastener system on wheel-rail dynamics characteristics

LI Xia, TANG Wei, DAI Jia-yu, SHENG Xiao-zhen, WANG An-bin

Abstract: More> To study the influence of frequency-dependent characteristics of double-layer nonlinear fastener system on wheel-rail dynamics characteristics, the dynamic stiffness of the fastener system within the frequency range of 100-1 000 Hz was obtained by hammering method. The finite element models of track structure and wheel-rail transient rolling contact were established. The frequency-dependent characteristics of the fastener system were considered in the models. Based on the track structure model, the influence of the frequency-dependent characteristics of the fastener system on the rail vibration characteristics was analyzed. The influence of the fastener system on the wheel-rail dynamics characteristics was analyzed by using the transient rolling contact model. Research results show that the dynamic stiffnesses of the fastener system in the vertical, lateral, and longitudinal directions generally increase with frequency. The frequency-dependent characteristics of the fastener system lead to more significant rail vibration at the frequency of above 340 Hz, and the frequency range corresponding to the peak rail vibration shifts towards higher frequencies. The changes in rail vibration response within the frequency range of 340-700 Hz are caused by the combined effects of frequency-dependent stiffness and damping, while the changes within the frequency range of 800-900 Hz are primarily influenced by the frequency-dependent stiffness of the fastener system. These changes in rail vibration response have a significant impact on the passing frequency of short-wavelength rail corrugation at above 340 Hz in small-radius curve sections. Therefore, considering the frequency-dependent characteristics of the fastener system is beneficial for improving the prediction accuracy of short-wavelength rail corrugation in this frequency range. The frequency-dependent stiffness of the fastener system has a relatively small impact on axle box vibration acceleration and wheel-rail forces, and the maximum value reduces by about 2%. However, the frequency-dependent damping of the fastener system has a significant impact on both. The frequency-dependent damping increases axle box vibration acceleration and wheel-rail forces within the frequency range of 100-470 Hz and reduces them within the frequency range of 470-1 000 Hz. The frequency-dependent characteristics of the fastener system have little effect on the maximum wheel-rail contact stress. When frequency-dependent stiffness is considered, the corresponding wheel-rail contact stress is slightly higher, and when frequency-dependent damping is considered, the corresponding wheel-rail contact stress is slightly lower.

2024, 24(6): 159-171. doi: 10.19818/j.cnki.1671-1637.2024.06.011

FullText HTML (22) PDF (12504KB) (8)

Magnetic force calculation method of permanent magnet electromagnetic hybrid suspension system considering magnetic saturation

SHEN Lu, ZHANG Li-wei, XIU San-mu, ZHANG Meng-lei, LIU Tao

Abstract: More> To improve the accuracy and efficiency of magnetic force calculation of permanent magnet electromagnetic hybrid suspension system, the advantages of fast calculation speed of analytical method and high calculation accuracy of finite element method were comprehensively considered. Based on the equivalent surface current method, the permanent magnet in the system was equivalent to an electromagnetic coil. Then, the pure electromagnetic coil equivalent model of permanent magnet electromagnetic hybrid suspension system was established. Based on the current-flux linkage expression of the inductance of electromagnetic coil, namely the nonlinear magnetic saturation model and virtual displacement energy balance method, the analytical expression of the magnetic force for the permanent magnet electromagnetic hybrid suspension system was derived. Based on the finite element simulation results, the parameter variable-air gap equation was fitted, and the magnetic force calculation method of permanent magnet electromagnetic hybrid suspension system considering the magnetic saturation was determined. Analysis results show that the nonlinear magnetic saturation model of the inductance of electromagnetic coil can effectively express the influence of magnetic saturation on magnetic force calculation. When the current flowing through the electromagnetic coil of permanent magnet electromagnetic hybrid suspension system is 0, 2.0, 8.5, and 17.0 A, the average deviations of the magnetic force calculation results obtained by traditional analytical calculation formula and traditional finite element simulation results are 32.87%, 20.02%, 29.05%, and 25.67%, respectively. The average deviations of the magnetic force calculation results obtained by magnetic force calculation method considering the magnetic saturation compared with the traditional finite element simulation results are 1.68%, 1.64%, 1.58%, and 1.37%, respectively, and the average deviations compared with the test results are 1.66%, 1.77%, 0.91%, and 1.35%, respectively. Compared with the traditional analytical calculation formula, the proposed method greatly improves the accuracy of magnetic force calculation. Compared with the traditional finite element simulation method, the amount of calculation tasks of the proposed method reduces greatly, and the calculation efficiency improves greatly.

2024, 24(6): 172-182. doi: 10.19818/j.cnki.1671-1637.2024.06.012

FullText HTML (11) PDF (6447KB) (7)

Short-term traffic flow velocity prediction model based on multi-channel data fusion of meteorological and traffic

MA Fei, YANG Zhi-jie, WANG Jiang-bo, SUN Qi-peng, BAO Bo, GUO Qing-yuan, HUANG Kai

Abstract: More> In order to improve the prediction accuracy of short-term traffic flow under the influence of meteorological and traffic factors, a short-time traffic flow velocity prediction model based on Gramian angular field-convolutional neural network-long short-term memory (GAF-CNN-LSTM) was proposed by considering the fusion of meteorological and traffic feature data such as temperature, humidity, and traffic congestion index. By utilizing the Gramian angular field, time series data of temperature, humidity, and historical traffic flow velocity were transformed into image data. RGB multi-channel color coding was employed to generate meteorological and traffic feature images, and the color superposition changes in the RGB muti-channel reflected the meteorological and traffic multi-factor feature data fusion. The transformed feature images were input into a convolutional neural network to extract fusion features of meteorological and traffic factors. An long short-term memory neural network was employed to capture the time series information of meteorological and traffic factors, thereby a short-term traffic flow velocity prediction model was constructed. Meteorological data, including temperature, humidity, and historical traffic flow velocity data from the Weiyang District in Xi'an were selected. According to the changing trend of temperature and humidity, two extreme scenarios of minimum humidity and minimum temperature with maximum humidity, and two conventional scenarios of non-extreme temperature and humidity were set to verify the model. Analysis results indicate that the GAF-CNN-LSTM model can consider the fusion of meteorological and traffic multi-factor feature data. Compared with moving average model, autoregressive integrated moving average model, prophet model, random forest model, and LSTM neural network model, the mean square error, root mean square error, and mean absolute percentage error of GAF-CNN-LSTM model reduce by 0.044 6, 0.142 4, and 12.1%, respectively, while the determination coefficient shows an average improvement of 31.05%. The GAF-CNN-LSTM model exhibits the highest prediction accuracy, and the results can provide a more accurate decision-making basis for urban transportation governance.

2024, 24(6): 183-196. doi: 10.19818/j.cnki.1671-1637.2024.06.013

FullText HTML (20) PDF (9472KB) (11)

Joint optimization model of tram departure plan and multi-modal and multi-path traffic green waves

WANG Hao, XIE Ning

Abstract: More> Addressing the issue of low traffic efficiency of trams and incoordination between trams and motor vehicles, a joint optimization model of tram departure plan and multi-modal and multi-path traffic green waves was proposed. The relationship between the frequency of signal priority control activation at intersections and traffic signal control schemes of trams and tram departure intervals was analyzed, and traffic signal cycles at intersections and tram departure intervals were coordinated to minimize the priority control frequency. A multi-modal and multi-path traffic signal coordinated control model was established with the optimization objective of maximizing the green wave bandwidth for road motor vehicles. Based on the model, the offsets, phase sequences, and travel times of road traffic and trams can be optimized, and the multi-path segmented green waves for road motor vehicles and multi-path green waves for trams can be generated as well. The model was verified by a case study and compared with other two models. Research results indicate that compared to the Multiband model and other time table and signal control joint optimization model, the proposed method coordinates the traffic signal cycles and tram departure intervals, and reduces the number of priority controls by 6 and 4, respectively, significantly reducing the impact of priority control on road motor vehicles. The model also optimizes the traffic signal control schemes and provides multi-path green waves for both trams and road motor vehicles, with the minimum bandwidth of motor vehicle green wave no less than 25 s. Compared to the other two models, the average speed of the proposed method improves by 29.36% and 27.85%, the average delay reduces by 42.24% and 35.03%, and the average total travel time reduces by 30.75% and 28.02%, respectively. Considering the improvement of multi-modal traffic network efficiency of tram priority control, the proposed model reduces the per capita delay by more than 17.60%.

2024, 24(6): 197-211. doi: 10.19818/j.cnki.1671-1637.2024.06.014

FullText HTML (16) PDF (4256KB) (9)

Mixed platoon control method considering traffic oscillation mitigation

DONG Chang-yin, XIONG Zhuo-zhi, LI Ni, WANG Feng, ZHANG Jia-rui, WANG Hao

Abstract: More> An information interaction form between human-driven vehicle (HDV) and connected automated vehicle (CAV) in a mixed platoon was established, based on which a feedback feedforward controller was designed for CAVs. Besides, an optimization method targeting the control parameters of CAVs was proposed to mitigate traffic oscillations. A circular closed system with a mixed platoon was formulated, where CAVs can transfer information through vehicle-to-vehicle communication spaced by HDVs. Based on the modelling of HDV car-following behaviors and their uncertainty, a third-order vehicle dynamics model, constant time gap policy, and feedback feedforward controller were used to design the control strategy for CAVs. A new generation of simulation data and fast Fourier transform were used to analyze the predominant frequency range of speed fluctuation of HDVs, and an index of the suppression degree of CAVs to speed fluctuation was constructed. By considering the uncertainty of HDV behaviors, an objective function was constructed for this frequency range to simultaneously optimize the string stability and suppress speed variation. Based on the real vehicle trajectory data, the control method was evaluated with simulations from multiple dimensions under the scenarios of different market penetration rates and spatial distributions of CAVs. Analysis results show that compared with reference strategies, the average acceleration variation of the individual CAV reduces by 10.9%-14.1%, and the maximum speed variation reduces by 7.8%-10.8%. The deceleration rate to avoid the crash reduces by 1.8%-21.6%, and the fuel consumption reduces by 2.9%-3.9%. For the whole mixed platoon, the comfort, stability, safety, and energy saving all improve when CAVs are uniformly distributed. The improvement effect is significant under the medium market penetration rate of 30%-60%. So, the control method can effectively suppress speed variation and substantially improve the ability of CAVs to mitigate traffic oscillations.

2024, 24(6): 212-229. doi: 10.19818/j.cnki.1671-1637.2024.06.015

FullText HTML (9) PDF (17997KB) (10)

System design of heterogeneous vehicle platoon based on multi-delay proportional-retarded controller

ZHU Xu, ZHANG Lin-hu, YAN Mao-de

Abstract: More> To effectively utilize delays to enhance system response speed and other control performances, a proportional-retarded (PR) controller was designed for the heterogeneous vehicle platoon by introducing multi-delays. The characteristic equations of each subsystem of the vehicle platoon were analyzed by applying the implicit function theorem and Cauchy-Riemann equation, and the right-most pole assignment method was proposed for these subsystems. The available assignment ranges of the right-most poles were determined, and the guidelines for designing the controller gains and delays were given. To gradually increase the stability of the subsystems that corresponded to the follower vehicles, the poles of subsystems were separated reasonably, and a right-most pole assignment method was proposed for the entire vehicle platoon. Moreover, sufficient conditions for string stability were derived. Simulation results indicate that the designed multi-delay PR controller and the proposed right-most pole assignment method can accurately assign the pole of the heterogeneous vehicle platoon at the desired location, and both the internal stability and string stability of the vehicles are ensured. Moreover, a smaller right-most pole indicates a faster system response speed. Compared to a single-delay PR controller, the right-most poles of subsystems are flexibly separated via this method, the system setting time reduces by 2.45%. Compared with traditional proportional-derivative (PD) controllers, the PR controller shows significant improvements in ride comfort and fuel economy, with enhancement ranging from 2-4 orders of magnitude. Under the disturbances of the same frequency, the amplitude ratios of the acceleration and control input for both the PR and PD controllers remain consistently below 1/3, dropping to 0.031 when the frequency of the disturbances is 9 Hz. It can be seen that the designed multi-delay PR controller has significant advantages in response speed, pole separation, and disturbance suppression.

2024, 24(6): 230-242. doi: 10.19818/j.cnki.1671-1637.2024.06.016

FullText HTML (14) PDF (5778KB) (7)

Homogeneous platoon speed planning and following control in front vehicle cut-in and cut-out conditions

YANG Wei, SUN Xue, SI Yu, HAN Yi, CAI Yao

Abstract: More> To solve the problems that the front vehicle cut-in and cut-out may lead to a collision risk, as well as the low efficiency and poor stability of following control during constant speed cruise of the homogeneous platoon, a homogeneous platoon speed planning and following control model in front vehicle cut-in and cut-out conditions was proposed. Based on the convolutional neural network-gated recurrent unit (CNN-GRU) hybrid network, a trajectory prediction model was established to predict the lane change trajectory of front vehicle within a future time domain. By constructing the relationship between longitudinal displacement and time, the front vehicle cut-in or cut-out state was determined. The speed following model of following vehicles was built, and the overall speed planning of the platoon was carried out. A fuzzy linear upper controller was established to output the desired acceleration according to the speed difference meeting the driving scenario requirements. Combining the longitudinal inverse dynamics model and fuzzy proportional integral derivative (PID) control, a lower controller was established to convert the desired acceleration into the driving torque or braking pressure. Thus, the direct control of the vehicle was achieved. To further verify the effectiveness of the proposed control model, a PreScan/CarSIM/SIMULINK co-simulation platform was built, simulation conditions for the adjacent front vehicle cut-in and cut-out were designed, and the proposed control model was compared with the adaptive cruise control (ACC) scheme. Research results show that after applying the proposed control model, no rear-end coallision occurs in platoons for all information flow topologies, and the maximum spacing difference, global average speed tracking error and local average speed tracking error reduce at least 55.9%, 40.6% and 42.0%, respectively. Therefore, the proposed control model can not only avoid potential rear-end collisions in front vehicle cut-in and cut-out conditions, but also can effectively shorten the maximum gap between platoon vehicles, reduce the speed tracking error of the platoon, and help to improve the following efficiency and driving stability of the platoon.

2024, 24(6): 243-258. doi: 10.19818/j.cnki.1671-1637.2024.06.017

FullText HTML (25) PDF (10703KB) (10)

Trajectory tracking control of intelligent vehicles based on deep reinforcement learning and rolling horizon optimization

XIE Xian-yi, ZHAO Xin, JIN Li-sheng, GUO Bai-cang, LI Ke-qiang

Abstract: More> To improve the generalization of trajectory tracking strategies for intelligent vehicles trained by deep reinforcement learning, a method of trajectory tracking control of intelligent vehicles based on rolling optimization and twin delayed deep deterministic policy gradient (ROTD3) was proposed to address the issue of poor trajectory tracking performance under different speed conditions when the reinforcement learning models were trained at a single speed. The trajectory tracking model was trained by a twin delayed deep deterministic policy gradient (TD3) deep reinforcement learning with a fixed speed tracking double lane change trajectory. Parameters of TD3 model were adjusted to obtain a strategy that satisfied the required trajectory tracking accuracy and achieved rapid convergence. Based on the trained TD3 model and the idea of model predictive control (MPC), a framework integrating ROTD3 was constructed. In the prediction horizon, the front-wheel steering angle output by the TD3 model was used for prediction. The lateral deviation and heading deviation in the course of trajectory tracking were optimized in the rolling horizon, and the control horizon sequence in the form of front-wheel steering angle increment of vehicles was solved by the quadratic programming method. The first control increment in the control horizon was added to the TD3 control quantity. This sum was used as the front-wheel steering angle control ourput. The ROTD3, TD3, and MPC were compared through the simulation experiments for trajectory tracking. Research results show that the ROTD3 achieves higher trajectory tracking accuracy. During the double lane change trajectory tracking at a longitudinal speed of 20 m·s^-1, the mean absolute lateral deviation of ROTD3 reduces by 83.52% compared with TD3, and reduces 91.02% compared with MPC. When tracking a snake-like trajectory, the results of ROTD3 are consistent with those of double lane change simulation. When the front-wheel steering angle output by the TD3 model results in large tracking deviations, the front-wheel steering angle increment obtained through the rolling horizon optimization effectively compensates for these deviations. The ROTD3 framework significantly improves the vehicle trajectory tracking performance under various conditions and effectively enhances the generalization and applicability of TD3 reinforcement learning trajectory tracking strategy.

2024, 24(6): 259-272. doi: 10.19818/j.cnki.1671-1637.2024.06.018

FullText HTML (21) PDF (11434KB) (11)

Train-induced environmental vibration analysis of large-chassis and multi-tower metro depots

ZHAO Mi, LU Cheng-lin, CAO Sheng-tao, WANG Wei-ning, SHEN Jia-xu, DU Xiu-li

Abstract: More> To investigate the vibration response patterns at large-chassis and multiple-tower metro depots under the influence of train-induced vibration load, a refined numerical analysis model and efficient analysis techniques were employed. Based on the substructure method, the generation and propagation of vibrations were decomposed into two substructures. The analytical expression for wheel-rail forces in a half-train model was derived based on the simplified calculation method of half-bogie train load. A refined finite element model for the track-soil-building system was established. By applying the wheel-rail force load to the finite element model, the dynamic response of the system was solved, and the high-efficiency analysis was achieved through the multi-graphics processing unit (GPU)-heterogeneous parallel explicit algorithm. By taking a specific metro depot as the research subject, the accuracy of numerical simulation was validated with the on-site measurement data. The efficiency of multi-GPU heterogeneous parallel explicit algorithm was confirmed by comparing the solution times of different algorithms applied to the same model. On this basis, the propagation patterns of vibrations within the operation depot and over-track buildings were highlighted. Analysis results indicate that within the operation depot, the vibrations generated by train operation are predominantly characterized by medium to low frequencies. The dominant frequency band of the source strength concentrates on 10-50 Hz, with a peak frequency near 30 Hz. The vibration level at the column base of the operation depot decreases linearly with the increase in the horizontal distance from the vibration source. However, due to the slight constraint exerted by the structure on perimeter columns compared with others, the vibration response amplifies on the outmost perimeter columns of the operation depot. Given that trains pass directly between the supporting columns at the lowest structural level, the vibration propagation path is short with weak attenuation, leading to a richer vibration spectrum in over-track buildings, where the floor vibration advantageously concentrates on the frequency band of 20.0-31.5 Hz. A coherent vibration pattern of the over-track building floor is observed at low frequencies, with the maximum Z-vibration level at mid-span near the top floor. The floor vibration intensity is influenced by both its natural frequency and the spectral characteristics of source strength. Resonance is more likely to occur when the natural frequency of floor falls within the predominant vibration frequency band of source strength.

2024, 24(6): 273-285. doi: 10.19818/j.cnki.1671-1637.2024.06.019

FullText HTML (14) PDF (9263KB) (5)

Evaluation of eye-catching effect of highway tunnel entrance zones based on factor analysis and data envelopment analysis

HAN Lei, DU Zhi-gang, HE Shi-ming, MA Ao-jun

Abstract: More> To comprehensively evaluate the impact of visual attractions in highway tunnel entrance zones on drivers' eye-catching effect, 30 participants were recruited for a naturalistic driving experiment. Eye movement and electrocardiogram (ECG) data were collected under different visual attraction conditions in the highway tunnel entrance zones. Sensitive evaluation indicators for the eye-catching effect were selected based on factor analysis. A data envelopment analysis (DEA) model was constructed to identify and explore the characteristics and mechanisms of how visual attraction conditions in highway tunnel entrance zones influence drivers' eye-catching effect. Analysis results show that the sensitive indicators for drivers' eye-catching effect in terms of visual characteristics are fixation duration(FD1), pupil diameter(PD), saccade duration(SD), and saccade range(SR), while the sensitive indicators for ECG characteristics are heart rate (HR), ration of low-frequency to high-frequency ratio (R), sample entropy (SampEn), and fractal dimension (FD2). Drivers' eye-catching effect is significantly affected by different visual attraction conditions in highway tunnel entrance zones. Specifically, under warning sign conditions, drivers exhibit significantly increase in average FD1 (530.97±37.03 ms), PD (4.56±0.46 mm), and SD (32.89±3.14 ms), along with a significant decrease in SR (4.77°±1.27°). Concurrently, a significant rise can be seen in HR (96.64±9.23 beates per minute), R (4.17±0.98), and FD2 (1.87±0.17), while a decline is found in SampEn (1.84±0.24). These findings suggest that under warning sign conditions, drivers' perception and processing of visual information are impaired, with higher visual cognitive load and psychological stress, leading to greater instability in their psychological state. Different visual attraction conditions have a significant impact on the overall efficiency of drivers' eye-catching effect. The mean efficiency is the highest (0.987) in a scenario with no significant visual attraction and the lowest (0.928) in a warning sign scenario, with significant differences observed among various scenarios. Saccade range and HR are the variables most affected by drivers' eye-catching effect. The warning sign has the most negative impact on drivers' visual characteristics and psychological load levels. The research findings provide valuable insights for optimizing the visual environment design of highway tunnel entrance zones, contributing to effective management and control of driving risks in such zones.

2024, 24(6): 286-298. doi: 10.19818/j.cnki.1671-1637.2024.06.020

FullText HTML (12) PDF (11657KB) (7)

Spatiotemporal characteristics and severity modelling of electric vehicle-pedestrian collision accidents

ZHUANG Yan, DONG Chun-jiao, LI Peng-hui, ZHENG Rui

Abstract: More> To investigate the spatiotemporal distribution pattern of electric vehicle-pedestrian collision accidents, the indicators of clustering strength and accident severity were incorporated into the traditional kernel density estimation method. A spatiotemporal kernel density estimation approach for traffic accident-prone locations was built. The random parameter Logit model based on mean heterogeneity was developed to model the severity of electric vehicle-pedestrian collision accident by considering the potential random variables and heterogeneity in accident variables. Key factors affecting the severity of electric vehicle-pedestrian collision accidents were identified, and the marginal utility of each significant variable was calculated to quantify its impact on the severity of electric vehicle-pedestrian collision accidents. Research results show that the electric/fuel vehicle-pedestrian collision accidents are predominantly located in urban center areas, whereas the distribution of electric vehicle-pedestrian collision accidents is relatively more concentrated. Moreover, the number and severity of electric vehicle-pedestrian collision accidents are higher on weekdays and daytime. Electric vehicle driver age ranging from 40 to 60 years old and accident occurring time between 18:00 and 6:00 the next day are identified as random variables. The utility value of driver age follows a normal distribution with standard deviation of -0.582 and variance of 0.926², and has mean heterogeneity in variables of road function level (trunk road) and vehicle type (truck). The utility value of accident occurring time follows a normal distribution with standard deviation of -0.313 and variance of 0.518², and there is mean heterogeneity in variable of weather (haze). Additionally, the factors such as male drivers, middle-aged and young drivers, and weekday travel are found to contribute to the varying degrees of severity in electric vehicle-pedestrain collision accidents. The Akaike information criterion of the random parameter Logit model considering mean heterogeneity decreases by 26 compared with the random parameter Logit model, and the logarithmic likelihood value increases by 94.272. The introduction of mean heterogeneity improves the fit goodness of severity model for electric vehicle-pedestrian collision accidents based on the random parameter Logit model, providing a more scientific and comprehensive explanation of causal principles behind electric vehicle-pedestrian collision accidents.

2024, 24(6): 299-310. doi: 10.19818/j.cnki.1671-1637.2024.06.021

FullText HTML (14) PDF (7380KB) (9)

User Center

Journal Info

Collection of Source JournalsMore>

Current Issue
2024, Volume 24, Issue 6

Search Type

Search Journal

LinksMore>