交通运输工程学报

Cover and Contents of Vol.23, No.3, 2023

Cover Contents

Review on aerodynamic noise analysis of high-speed train pantographs

LIU Hai-tao, WANG Wen-yu, ZHOU Xin, ZHANG Chang-liang, XIAO Qian

Abstract: In order to deeply and comprehensively understand the research status of aerodynamic noise of high-speed train pantographs and elucidate its mechanisms and laws, the research on the aerodynamic noise in China and abroad in recent years was summarized. Its development histories in China, Japan, Germany, and France were outlined. The sources of aerodynamic noise of pantographs, the characteristics of radiated aerodynamic noise, and the research methods for the aerodynamic noise were analyzed. The generation mechanisms and suppression methods of the aerodynamic noise were explored, and the main achievements obtained in the current research were summarized. Analysis results indicate that, as an important current collection device on the top of the train, the pantograph consists of multiple rods, and it generates significant tonal noise in high-speed airflow and is one of the main sources of environmental noise pollution in high-speed trains. The main sources of aerodynamic noise of high-speed train pantographs are distributed on the windward sides of components such as the pantograph head, knuckle, insulators, and base frame. The methods for studying the aerodynamic noise of pantographs include actual train test, wind tunnel experiment, and numerical simulation. The addition of auxiliary devices, such as the wind deflector, jet flow, and plasma actuator, can effectively control the aerodynamic noise, but these methods increase the complexity of the system. According to the principle of bionics, modifying the surface microstructure of rods can significantly suppress the generation of turbulent vortices of pantographs, thus greatly reducing the aerodynamic noise. Optimizing the cross-sectional shape and spatial structure design of rods can reduce the generation of drag and turbulent vortices, thereby effectively controlling aerodynamic noise. So, various approaches can be employed to reduce the aerodynamic noise of pantographs. However, the feasibility of engineering implementation, the coupling relationship between aerodynamic noise and aerodynamic resistance, as well as the contact stability between the pantograph and catenary, still need to be further investigated.More>

2023, 23(3): 1-22. doi: 10.19818/j.cnki.1671-1637.2023.03.001

FullText HTML (388) PDF (22868KB) (220)

Review on aircraft towing taxi technologies

SUN Yan-kun, ZHANG Wei, YANG Xiong-wei, LIU Ji-hao, ZHU Heng-jia, LIU Yan-xi, QIN Jia-hao

Abstract: The essence of aircraft towing taxi mode was studied. The connotation and driving factors of towing taxi were proposed, and six key issues to realize aircraft towing taxi were put forward, including kinematics and environmental load analysis during towing taxi, multibody system dynamics of towing taxi, structural mechanical responses of the nose landing gear of aircraft, motion perception and control of the system, towing taxi equipment and physical test platform, and impact of towing taxi on airport operation rules. In addition, the research status and development trend of the six key issues were analyzed, and the technical difficulties and challenges faced in implementing the new mode were explored. Analysis results show that compared with traditional towing, the kinematics calculation of towing taxi requires higher precision and real-time performance, a high-precision algorithm of system motion should be established to realize the safety estimation of towing taxi, and the coupling effect of various factors such as the towing taxi system motion status of high-speed and heavy-duty aircraft, airport road conditions, and control inputs, may lead to limit phenomena, such as lateral instability and detachment. The research on the dynamic behavior of the system needs to consider the nonlinear characteristics of the structure and system. In order to ensure the safety of nose landing gear subjected to longitudinal traction and vertical vibration coupling conditions under long-distance towing taxi, it is necessary to study its structural dynamic response and ultimate working condition behavior mechanism. Man-machine interactive mode and pilot perception control method need to be analyzed to achieve the precise control of towing taxi movements by pilots, and tests are required for the verification. The using time of aircraft engines during ground operation can be shortened by towing taxi technology, and fuel consumption and carbon emissions can be reduced, but the complexity of airport surface operations will increase. Therefore, a new scene operation management rules in airport flight area system should be established.More>

2023, 23(3): 23-43. doi: 10.19818/j.cnki.1671-1637.2023.03.002

FullText HTML (415) PDF (2500KB) (148)

Review on machine learning-based traffic flow prediction methods

YAO Jun-feng, HE Rui, SHI Tong-tong, WANG Ping, ZHAO Xiang-mo

Abstract: The research status and development trend of macro traffic flow prediction of designated road sections and regional road network at home and abroad were analyzed by literature review, expert interview, and experimental scenario construction. Local section traffic flow prediction methods were summarized, including traditional machine learning, recurrent neural networks, and hybrid models. The characteristics of convolutional neural networks, graph neural networks, and fusion multi-factor networks were discussed.The principles, advantages, limitations, and application scenarios of the methods were explained. The types of existing scenario traffic datasets and the mainstream traffic datasets at home and abroad were summarized from the perspectives of sampling periods and collecting methods. Analysis results show that recurrent neural networks can effectively obtain the historical laws of traffic data, but there are some problems such as gradient explosion, high computational complexity, and poor accuracy of long-time prediction. Graph neural networks introduce graph structures for road network topological connection relationships, which has obvious advantage in considering the spatiotemporal correlation of road network and traffic flow data. Fusion multi-factor methods fully consider the influence of internal and external factors such as weather, roads, and accidents, effectively improving the real-time performance and robustness of traffic flow prediction. The improvements of traffic flow prediction methods have limitations due to the difficult traffic data collection and external factor influence quantification, as well as the poor interpretability of machine learning methods. The future research should start from two aspects of starting the efficient mining of traffic information and the perfection of graph convolution methods, broaden the application of graph structures in the traffic field, and consider non-constant traffic scenarios. So as to further reveal the inherent laws of traffic data, develop more accurate and efficient traffic flow prediction methods, and promote the application of traffic flow prediction in industry.More>

2023, 23(3): 44-67. doi: 10.19818/j.cnki.1671-1637.2023.03.003

FullText HTML (888) PDF (11856KB) (1052)

Screening method of suspected irregularity for medium- and low-speed maglev tracks based on Box-Whisker plot

WU Jun, LI Hong-lu, ZHANG Yu-xin, ZHANG Yun-zhou, TANG Jun-yuan

Abstract: In order to reduce the workload of medium- and low-speed maglev track detection and the time occupied by track detection equipment and improve the efficiency of track irregularity detection, a method for screening suspected track irregularities by collecting the relevant data of suspension system through the driving recorder was proposed. According to the basic idea that the abnormal irregularity of the track line would cause obvious mutation anomalies to the suspension gap, vertical acceleration of the suspension electromagnet, and electromagnet current, based on the Box-Whitker plot, the screening threshold was set up, and the abnormal screening of the ternary threshold was carried out by using the data of suspension gap, vertical acceleration of the suspension electromagnet, and electromagnet current, and the data that might be missed were screened again based on the monadic and binary thresholds. The above results were used to determine the abnormal level of the track section, and the abnormal sections that had been screened many times were judged as sections with suspected irregularity. To further improve the accuracy of the screening results of abnormal sections with irregularity, according to the data of different suspension control points of the same carriage passing through the same abnormal section with irregularity, the idea of abnormal repeatability should be reflected, and the data of multiple suspension control points recorded by the driving recorder were be fused in terms of the screening results, so as to comprehensively judge the suspected irregularity of the section. On this basis, the proposed method was applied to analyze the data of 10 suspension control systems on the left side of M carriage on Changsha Maglev Line. Research results show that the proposed method can screen out suspected abnormal sections so that the existing whole-section irregularity detection method can be transformed into a method using track inspection equipment to detect suspected abnormal sections in a targeted manner. The maintenance time of line detection can reduce by about 20%.More>

2023, 23(3): 68-76. doi: 10.19818/j.cnki.1671-1637.2023.03.004

FullText HTML (240) PDF (3786KB) (59)

Arc distribution in pantograph-catenary contact based on double-pantograph current collection dynamics

ZHOU Hong-yi, LIU Zhi-gang, XIONG Jia-ming, XU Zhao, DENG Yun-chuan

Abstract: In view of the arc problem during double-pantograph current collection, the microstructure of the pantograph-catenary contact surface was analyzed based on the classical electrical contact theory, and the mechanism of arc occurrence in the pantograph-catenary contact state was discussed. The probability algorithm of arc occurrence in the pantograph-catenary contact was proposed, and the dynamics model of the double-pantograph coupling system was analyzed by considering the fluctuation and propagation laws of the catenary under the double-pantograph current collection. A finite element model of the catenary and a multi-body dynamics model of the pantograph were established to obtain the contact forces on the leading and trailing pantographs under different static lifting forces of trailing pantograph. According to the contact arc probability algorithm and the obtained contact forces, the probabilities of contact arc occurrence on the leading and trailing pantographs under multiple working conditions were calculated, the distribution law of the contact arc under the corresponding contact force was analyzed, and the measures to reduce the probability of contact arc occurrence were proposed. Research results show that the contact arc occurrence probability on the leading pantograph is much smaller than that on the trailing pantograph, its average value is only 32% of the latter average value. The change of static lifting force of the trailing pantograph has little impact on the contact arc occurrence probability on the leading pantograph. The contact arc occurrence probability on the leading pantograph fluctuates between 1.5×10^-3 and 5.0×10^-3 under four working conditions, and there is no obvious change rule. The contact arc occurrence probability on the tailing pantograph decreases as the static lifting force of the tailing pantograph increases. The average contact arc occurrence probability on the trailing pantograph reduces by 42% as the lifting force increases from 55 N to 85 N. There is no obvious law of contact arc occurrence on the leading pantograph with the distribution of contact force, and the higher static lifting force of the tailing pantograph makes the distribution more uniform. The distribution of the contact arc occurrence probability on the tailing pantograph decreases with the increase of the contact force, and the probability mainly distributes in the low-value range of the contact force. Controlling the contact force within the range of 70-80 N and greater than 170 N can effectively suppress the arc occurrence on the leading pantograph, and controlling the contact pressure more than 70 N can effectively suppress the arc occurrence on the trailing pantograph.More>

2023, 23(3): 77-87. doi: 10.19818/j.cnki.1671-1637.2023.03.005

FullText HTML (163) PDF (6483KB) (74)

Estimation method of corrugation state based on wheel-rail vertical force

NIU Liu-bin, HU Xiao-yi, YANG Fei, QIANG Wei-le

Abstract: In order to find out the mapping relationship between wheel-rail vertical force and corrugation state and to evaluate quantitatively the severity of the corrugation by using wheel-rail force inspection data, the typical parameters from CRTS Ⅱ high-speed railway and electric multiple units in service in China were employed to construct a three-dimensional wheel-rail dynamics finite element model. The characteristics of the irregularities on the rail surface at the corrugation section were refined, and the simulation accuracy of the constructed model was verified by the time-frequency data of the measured wheel-rail vertical force from the high-speed comprehensive inspection car at the corrugation section in the high-speed railway. On this basis, the wheel-rail vertical forces excited by the corrugation with a wavelength between 40 mm and 180 mm at the running speed of 300 km·h^-1 were simulated, and their distribution characteristics in time-frequency domain were analyzed. The change rate of rail surface irregularity was introduced to characterize the changing characteristics of the corrugation along the longitudinal direction of the rail. The nonlinear least square method and rational equation were used to fit the functional relationship between the large values of the wheel-rail vertical force and the change rates of rail surface irregularity under different wavelengths. The influence of rail vibration mode of the Pinned-Pinned natural resonant frequency and its half-value on the fitting parameter curves was analyzed. A method based on the wheel-rail vertical force was derived to estimate the valley depth of the corrugation. The valley depth estimation method was tentatively tested in the high-speed railway to evaluate the corrugation, 32 sets of corrugation sections were found, and the measured and estimated valley depths at the sections were compared. Analysis results show that the correlation coefficient between the estimated and measured valley depths is 0.97, so they have a high linear correlation. The root mean square error between the estimated and measured valley depths is approximately 0.01 mm when the estimated valley depth is greater than 0.08 mm, and the misjudgment rate in making decisions on rail grinding based on the estimated valley depth is approximately 6.25%, indicating that the estimation method of valley depth has good applicability in actual high-speed railway.More>

2023, 23(3): 88-102. doi: 10.19818/j.cnki.1671-1637.2023.03.006

FullText HTML (339) PDF (19016KB) (100)

Effects of glue injection and reinforcement anchor installation on mechanical deformation of longitudinally connected slab-type ballastless tracks

LI Yang, CHEN Jin-jie, WANG Jian-xi, SHI Xian-feng, CAI Jia-sheng

Abstract: For the mechanical deformation of longitudinally connected slab-type ballastless tracks under high-temperature load after maintenance measures were taken, the nonlinear constitutive relationship of bond stress-displacement at interfaces between track slabs and mortar layers and the nonlinear constitutive relationship of stress-slip in reinforcement anchor structure were considered, a finite element model for mechanical behavior analysis of longitudinally connected slab-type ballastless tracks was built, the nonlinear high-temperature load was applied to the model, and the influence laws of glue injection, reinforcement anchor installation, and 2 measures jointly used on the mechanical deformation and structural damage were compared and analyzed. Analysis results show that for the basic working condition with a 0.2 m interface seam on both sides of the track slab and a damaged wide and narrow joint, under the maintenance measures of glue injection, reinforcement anchor installation, and 2 measures jointly used, the maximum values of the interface seam at the end of the track slab adjacent to the damaged joint are 63%, 20%, and 18% of those before taking maintenance measures, and the maximum values of the compression damage of the damaged joint concrete are 51.0%, 6.8%, and 5.5% of those before taking maintenance measures. For the longitudinally connected slab-type ballastless track with a poor structural state of wide and narrow joints, the maintenance effect of reinforcement anchor installation is much better than that of glue injection, and the effect of using the two measures jointly is better. The effects of glue injection on limiting the vertical displacement at the end of longitudinally connected slab-type ballastless tracks, interface damage, and compression damage of wide and narrow joints increase with the increasing depth of glue injection. When only glue injection is taken, in order to reach the amplitude range of interface damage at the end of the track slab adjacent to the damaged joint when 2 measures are jointly used, the number of glue injection slabs shall not be less than 2, and the depth of glue injection on both sides shall not be less than 0.9 m. It is recommended to fully ensure the glue injection maintenance area when the glue injection is taken only.More>

2023, 23(3): 103-113. doi: 10.19818/j.cnki.1671-1637.2023.03.007

FullText HTML (232) PDF (7638KB) (59)

Number selection and structural optimization of 400 km·h^-1 high-speed turnout

WANG Pu, WANG Shu-guo, WANG Meng, ZHAO Zhen-hua, SI Dao-lin, MA Si-yuan, SUN Zhao-liang

Abstract: According to the demand for further improving the operating speed of high-speed railways in China, the number selection and structural optimization methods of a 400 km·h^-1 high-speed turnout were studied systematically. Based on the vehicle-turnout coupling dynamics simulation, the correlation between turnout number and ride comfort was analyzed, and the suggestions for number selection were given. The influence of turnout linetype on the dynamics performance was studied, the principle of utilizing existing turnout sleepers was considered, and a linetype optimization scheme was proposed. A prototype test platform for the switch rail conversion of the No.18 KEZHUANXIAN turnout was established. The influencing factors and mechanisms of insufficient switch rail displacement were studied, and a control method of the insufficient displacement was proposed. A calculation model of track stiffness in the turnout area was established based on the finite element theory. According to on-site measured data of rail dynamic displacement in the turnout area, a target value and a homogenization scheme for track stiffness were proposed. Research results indicate that under the condition of the existing station layout plan, it is recommended to choose the No.18 turnout as the 400 km·h^-1 high-speed turnout. The mutual distance of the 400 km·h^-1 high-speed turnout can be increased to 28 mm, which can significantly improve the wear resistance and service life of the switch rail and can utilize the existing turnout sleepers. By comprehensively considering the system matching design, manufacturing process, combination requirements of track maintenance and electrical divisions, and other factors, it is recommended to reduce the distance from the third traction point to the fixed end by 600 mm, which can reduce the insufficient switch rail displacement. At the same time, the minimum flangeway and the traction force of the third traction point meet the code requirements. In addition, the structural design method of high-speed turnout crossings should be improved, in which the elastic deformation state of the point rail should be determined according to the actual forces, and the connectors and point rail conversion should be designed on this basis. It is also recommended that the target stiffness of the turnout area should be reduced from 25±5 kN·mm^-1 to 23±3 kN·mm^-1.More>

2023, 23(3): 114-126. doi: 10.19818/j.cnki.1671-1637.2023.03.008

FullText HTML (172) PDF (7297KB) (204)

Stiffness characteristics and life prediction of rail pads of subway damping fasteners

GAO Xiao-gang, FENG Qing-song, MA Yu-fei, WANG An-bin, SUN Hai-bo

Abstract: In order to study the stiffness sensitivity of rail pads of subway damping fasteners in service and their influence on line environmental vibration, the compression-type damping fasteners extracted from several operating lines of Nanjing Metro were taken as the research object, and the multi-environmental indoor comprehensive tests of the stiffness characteristics, fatigue characteristics at room temperature, and thermal accelerated fatigue aging characteristics of rail pads of the compression-type fasteners in service were carried out. Based on the test results, the correlations between service time and stiffness change of rail pads of new and old compression-type fasteners were compared and analyzed. The time-life characteristic curves of rail pads were obtained, and the life prediction model for the percentage change in stiffness and service time of the rail pads was proposed. Research results indicate that under the combined effect of periodic wheel-rail loads and the temperature, humidity, and alkaline environment of the line, the service stiffness of rail pads of the subway damping fasteners increases linearly with service time, and the elasticity of rail pads undergoes performance degradation. The stiffness curve trend of the new rail pads under thermal accelerated fatigue aging is basically consistent with the stiffness curve trend of the sampled rail pads in service. In other words, the thermal accelerated cyclic aging test of the rail pads can simulate or evolve on-site conditions such as the thermal mechanical cyclic load of the line. Based on the Arrhenius life-stress thermal accelerated aging model, the acceleration factors of rail pads under service stress and accelerated aging stress are 1.99 and 1.36, respectively, which can be used to predict the replacement cycle of rail pads of the damping fasteners.More>

2023, 23(3): 127-136. doi: 10.19818/j.cnki.1671-1637.2023.03.009

FullText HTML (224) PDF (11437KB) (100)

Integrated optimization method of vibration and sound radiation for urban rail wheel structure

WEN Yong-peng, LIU Yue-jie, ZHOU Yue, SHENG Xiao-zhen

Abstract: To reduce the wheel structure noise of urban rail vehicles, the double S-shaped spoke wheel in service was taken as the research object to build an optimization model of urban rail wheel structure noise considering the integration of vibration and sound radiation. A new noise reduction wheel profile with unequal thickness spokes from top to bottom was obtained, and an integrated optimization method of wheel structure vibration and sound radiation which took the rail wheel spoke area as the design area was proposed. The entire wheel spoke area was identified as the design domain, and the coding rules, selection rules, crossover rules, and variation rules were identified respectively. The optimization objective function of vibration and sound radiation converged gradually to evolve into a better noise reduction wheel profile. The vibration and sound radiation optimization design of the wheel was realized. The static strength, fatigue strength, and vibration sound radiation performance of the optimized wheel were calculated by the mature finite element tool, which further verified the effectiveness and reliability of the new structure noise optimization results of the double S-shaped spoke wheel. Research results show that the integrated optimization method of wheel structure vibration and sound radiation is suitable for the structural profile optimization of noise reduction wheels. The peak sound power level of the optimized wheel is 4.26 dB(A) lower than that of the original double S-shaped spoke wheel, and the noise reduction effect is obvious at sound power level peaks in the frequency range of 0-5 000 Hz. From the perspective of the structural characteristics of the spokes, the spokes of the optimized double S-shaped spoke wheel evolve from the basic equal-thickness spokes to unequal-thickness ones. The unequal-thickness characteristics of the wheel spokes are conducive to reducing the sound radiation of the wheel. Therefore, considering the economic and noise reduction performance of the wheel, it is suggested to take the profile wheel with unequal-thickness spokes as the noise reduction model of the wheel.More>

2023, 23(3): 137-147. doi: 10.19818/j.cnki.1671-1637.2023.03.010

FullText HTML (164) PDF (6642KB) (58)

Effect of lift airfoils on characteristics of slipstream and wake flow of high-speed trains

XIONG Xiao-hui, WANG Xin-ran, ZHANG Jie, WANG Kai-wen, CHENG Fan, LUO Zhang-jun

Abstract: A 1∶10 three-car CRH high-speed train model was taken as the research object to explore the drastic change of the flow field around the high-speed train caused by the installation of lift airfoils on the roof. An improved delayed detached eddy simulation (IDDES) method based on the two-equation turbulence model was adopted to analyze the development tendencies of the time-averaged and instantaneous slipstreams of two high-speed trains with and without lift airfoils. The distribution characteristics of instantaneous vortex structures in the wake region were discussed by a vortex identification method. The correlation between the peak slipstream velocity and unsteady characteristics of wake vortices was verified by the comparison of the slipstream distribution characteristics at different flow positions in the wake region and the movement laws of wake vortices. The power spectrum density curves of the velocity in the wake region were obtained by means of the spectral analysis. Research results show that due to the geometric structure of lift airfoils, the boundary layer separation on the train surface is intensified, and the thicknesses of the boundary layers on the roof and side surfaces of the train increase. The peak slipstream velocity is raised by the lift airfoils. Specifically, the maximum time-averaged slipstream velocities at the trackside and platform position increase by 1.556 and 1.327 times, respectively. It is delayed compared with the second peak position of the traditional train. Due to the continuous development and accumulation of wing-tip vortices downstream, the wake flow structure of the train with lift airfoils is manifested as a large-scale vortex pair mixed with a pair of more broken small vortices. Compared with the traditional train, the shear effect between the vortex and the ground is stronger, the time-averaged slipstream velocity of the wake flow of the train with lift airfoils is larger in the spanwise distribution but smaller in the vertical distribution. Moreover, there is a more obvious shear separation on the horizontal plane. Many small-scale broken vortices are incorporated in the wake of the train with lift airfoils, affecting the shedding frequency of vortices in the wake. As a result, compared with the traditional train, the train with lift airfoils has higher energy and slower vortex dissipation velocity.More>

2023, 23(3): 148-161. doi: 10.19818/j.cnki.1671-1637.2023.03.011

FullText HTML (236) PDF (26105KB) (51)

Health assessment method of traction motor bearing based on transfer learning and convolutional neural network

JIANG Ling-li, LI Shu-hui, LI Xue-jun, WANG Guang-bin, GAO Lian-bin

Abstract: To address the difficulties in acquiring labeled life-cycle vibration data and constructing health indicators that can reflect the bearing performance degradation trends in the health assessment of traction motor bearing, a health assessment method based on a transfer learning and convolutional neural network model was proposed for implementing the health assessment of traction motor bearing. The labeled bearing life-cycle data set was used as the source domain data by using the transfer learning, and the comprehensive test bench data were used as the target domain data to construct the data set. Undersampling and synthetic minority over-sampling techniques were used to expand and balance the life-cycle data set, and the abundant sample for convolutional neural networks training was obtained. The features describing the bearing degradation process were extracted in time domain and frequency domain. By using convolutional neural network and following the bathtub curve of bearing performance degradation, the health indicator was constructed by fusing the basic characteristics. Analysis results show that in the health assessment of traction motor bearing shaft current damage, the accuracy of the proposed health assessment method based on transfer learning and convolutional neural network is 98.17%. The accuracies of the methods constructing the health indicator according to linear, quadratic function and parabolic degradation law are 86.61%, 89.56% and 91.30%, respectively. Therefore, the proposed health assessment method has higher accuracy and better evaluation effect in the application. Moreover, the combination of expert knowledge and neural network learning knowledge reduces the fault characteristic dimension, solves the difficulty of health indicator construction, and realizes the health assessment of traction motor bearings through cross-device transfer learning.More>

2023, 23(3): 162-172. doi: 10.19818/j.cnki.1671-1637.2023.03.012

FullText HTML (199) PDF (12322KB) (83)

In-vehicle image technology for identifying faults of pantograph

DING Jian-ming, ZHOU Jing-yao, JIANG Hai-fan

Abstract: In view of the problem that the operation safety of the train was affected by pantograph faults, an in-vehicle image technology for identifying pantograph faults was proposed to detect the dropping, deformation and destruction of pantograph, the abnormal wear and notch of carbon contact strip, and deformation and loss faults of pantograph horns in real time. Based on the faster region-convolutional neural network (Faster R-CNN) target detection framework, a target detection model for locating the pantograph bow images was designed, and the residual network was used to replace the original convolutional network. The candidate region recommendation network was constructed by using the feature pyramid multi-scale prediction structure, so as to accurately and quickly locate the pantograph bow and detect the status. Based on the mask region-convolutional neural network (Mask R-CNN) instance segmentation framework, a pantograph bow image segmentation model was designed, and the network structure and feature map size of the detection head were redesigned to adapt to the slender and curved features of the pantograph, so as to accurately and quickly segment the pantograph bow image. In order to identify and locate faults more quickly in the segmented binary image, a rapid template matching strategy for the faults of pantograph horn and carbon contact strip was formulated according to the pantograph structure size and the position coordinates output by the image segmentation model. On this basis, detailed fault detection algorithms and procedures were compiled. Research results show that on the corresponding dataset, the average detection accuracy and average detection time per frame of the target detection model for positioning the pantograph bow images are 0.944 and 0.029 s, respectively. The average segmentation accuracy and average detection time per frame of the pantograph bow image segmentation model are 0.967 and 0.031 s, respectively. In addition, the detection accuracy and average detection time per frame of the template matching are 0.985 and 0.005 s, respectively. The average detection accuracy and average detection time per frame of the fault detection algorithm are 0.966 and 0.051 s, respectively. Thus, the proposed detection algorithm has high reliability and real-time performance.More>

2023, 23(3): 173-187. doi: 10.19818/j.cnki.1671-1637.2023.03.013

FullText HTML (578) PDF (14370KB) (101)

Multi-mode data augmentation and fault diagnosis of gearbox using improved ACGAN

SHAO Hai-dong, LI Wei, LIN Jian, MIN Zhi-shan

Abstract: To address the problems that the existing generative adversarial network (GAN) was difficult to efficiently generate multi-mode fault samples, and its training was unstable, an improved auxiliary classification GAN (ACGAN) was proposed for multi-mode data augmentation and intelligent fault diagnosis of the gearbox to ensure the safe operation of the vehicle. The independent classifier was introduced to construct a new ACGAN framework, and the compatibility between the classification accuracy and discriminant accuracy of classic ACGAN was improved. Wasserstein distances were used to define new adversarial loss functions with smooth properties, so as to overcome the disadvantages of the GAN, such as mode collapse and gradient vanishing. In order to improve the stability during the adversarial training process, the spectral normalization was used to replace the weight clipping to constrain the weight parameters of the discriminator. In order to verify the effectiveness and superiority of the improved ACGAN method, experimental analysis was performed on gearbox samples under six health conditions. Analysis results show that the fault samples generated by the improved ACGAN achieve better quality evaluation results at the data level and feature level, among which the evaluation index based on the structural similarity outperforms the comparison method by 0.249 3 on average, and the evaluation index based on the maximum mean difference outperforms the comparison method by 0.696 6 on average. The training process for the improved ACGAN is more stable. Its loss function has better convergence, and it has higher efficiency in multi-mode fault diagnosis scenarios with its training time reduced to 20% of the comparison method. In the case of missing fault samples, the generated fault samples of the improved ACGAN can effectively assist the training of intelligent fault diagnosis models based on deep learning, which can improve the diagnosis accuracy from 75.34% to 97.06%.More>

2023, 23(3): 188-197. doi: 10.19818/j.cnki.1671-1637.2023.03.014

FullText HTML (118) PDF (7744KB) (76)

Hub-and-spoke emergency rescue network planning

MA Chang-xi, SHI Chu-wei, DU Bo

Abstract: In order to achieve the rapid rescue of disaster-stricken cities and simultaneously reduce the construction cost of emergency rescue networks as much as possible, the multi-objective planning method of the three-level hub-and-spoke emergency rescue network was studied by taking the locations of emergency rescue stations and the layout of emergency rescue channels as the foothold. The multi-level structure of the hub-and-spoke network and the characteristics of connection relationship between emergency rescue stations were considered, the locations of three-level emergency rescue stations, the connection relationship between emergency rescue stations, and the level of emergency rescue channels were taken as decision variables, the minimum construction cost of emergency rescue stations at all levels and the minimum average network rescue time were taken as the two objective functions, a three-level hub-and-spoke emergency rescue network planning model was bulit. The niche Pareto genetic algorithm with a three-segment encoding structure for the three-level hub-and-spoke emergency rescue network planning model was designed by combining the characteristics of the decision variables. The emergency rescue network was modelled and solved based on the road network of 14 cities in Gansu Province. The effectiveness of the method was verified, and the optimization results were compared with the traditional three-level emergency rescue network. Research results show that the niche Pareto genetic algorithm with a three-segment encoding structure can effectively solve the hub-and-spoke emergency rescue network planning model. Compared with the optimal solution of the traditional emergency rescue network model, the selected scheme of the Pareto solution in the three-level hub-and-spoke emergency rescue network can reduce the construction cost of emergency rescue stations by 8.3%, and accelerate the average network rescue time by 3.5 h. The optimization results can dominate the optimal solution of the traditional emergency rescue network. So, the proposed three-level hub-and-spoke emergency rescue network planning method can take into account the intensive characteristics of hub-and-spoke networks and achieve shorter emergency rescue time.More>

2023, 23(3): 198-208. doi: 10.19818/j.cnki.1671-1637.2023.03.015

FullText HTML (321) PDF (2418KB) (107)

Resilience-based protection decision optimization for metro network under operational incidents

LU Qing-chang, LIU Peng, XU Biao, CUI Xin

Abstract: The protection decision optimization problem for metro networks was studied to alleviate the negative impacts triggered by operational incidents and improve the capability of metro networks to tackle these incidents. For the network resilience, the variation characteristics of the resilience curve and cumulative loss of the performance in the degradation and recovery of network performance were considered, and a two-layer optimization model for metro network protection decisions was constructed. The upper model was a stochastic integer programming model for identifying the optimal choice of the stations to be protected in the scenarios of uncertain operational incidents. The lower model was a user equilibrium assignment problem, and the variations in the queueing passenger flow and waiting time for recovery at the stations with limited capacity were prioritized to accurately estimate the delay time for passenger travel under operational incidents. The genetic algorithm and Frank-Wolfe algorithm were used to solve the upper and lower models, respectively. The metro network in the central area of Xi'an was taken as an example to verify and analyze the proposed models and algorithms. Analysis results show that the resilience-based protection decision is capable of reducing the loss of network performance by more than 50% by protecting 37.5% of the stations in the research region. It is superior to the vulnerability-based protection decision and the one without considering the substitution role of the bus network. Compared to the situation of the vulnerability-based one, the losses of network performance and passenger flow time reduce by 6.18% and 582 h, respectively, when half of the metro stations in the network are protected by the resilience-based protection decision. The protection priorities of more than two-thirds of stations in the metro network alter due to the substitution role of the bus network. For the same type of stations, their dependence on the substitution role of the bus network enhances with the increase in the passenger flow. The protection priority of a metro station is dependent mainly on the passing passenger flow and transportation capacity. A larger passenger flow is accompanied by a lower transportation capacity and higher protection priority of corresponding stations. The station type is also a factor determining the protection priority, especially for the stations with large passenger flows, and higher protection priorities are required for the non-transfer stations.More>

2023, 23(3): 209-220. doi: 10.19818/j.cnki.1671-1637.2023.03.016

FullText HTML (258) PDF (4975KB) (116)

Management and control method of dedicated lanes for mixed traffic flows with connected and automated vehicles

QIN Yan-yan, LUO Qin-zhong, HE Zheng-bing

Abstract: The vehicle types in the mixed traffic flow with connected and automated vehicles (CAV) and the headways under the car-following mode were analyzed, and the probability expressions of each headway mode were theoretically deduced according to the features of general mixed traffic flows, so as to mathematically describe the mixed traffic flow. In order to maximize the overall passing rate of mixed traffic flows, the setting conditions of a CAV dedicated lane in multi-lane mixed traffic flows and the optimal traffic flow distribution ratios of CAV traffic flows on the dedicated lane and the mixed lane after setting the dedicated lane were calculated. By extending the case of one CAV dedicated lane to the general case of dynamic management and control of multiple CAV dedicated lanes, an analysis method for dynamic management and control of dedicated lanes for the mixed traffic flows was constructed. Case analysis was used to demonstrate the effectiveness of the proposed management and control method of CAV dedicated lanes. Research results show that when traffic demand is 2 000 veh·h^-1, there is no need to set up CAV dedicated lanes in each CAV permeability stage. When the traffic demand is 3 000 veh·h^-1, CAV dedicated lanes should be set up at the CAV permeability stage of 0.2-0.4. When the traffic demand is 5 000 veh·h^-1, it is necessary to consider setting up CAV dedicated lanes in each CAV permeability stage. The optimal numbers of CAV dedicated lanes and the optimal distribution ratios of CAV traffic flows at different CAV permeability stages can be quantitatively calculated by the proposed CAV dedicated lane management and control method according to the conditions of traffic demand and the total number of lanes, and the critical CAV permeability range reflecting the setting conditions of CAV dedicated lanes can be affected by the traffic demand. The increase in the optimal number of dedicated lanes from the traffic demand attribute and the road space attribute can be promoted by the traffic demand and the total number of lanes, respectively, which is in line with the characteristics of dedicated lane management and control in multi-lane scenarios for mixed traffic flows.More>

2023, 23(3): 221-231. doi: 10.19818/j.cnki.1671-1637.2023.03.017

FullText HTML (219) PDF (4827KB) (205)

Car-following model and optimization strategy for connected and automated vehicles under mixed traffic environment

PENG Jia-li, SHANGGUAN Wei, CHAI Lin-guo, QIU Wei-zhi

Abstract: To improve the driving efficiency and traffic flow stability of vehicles in the mixed traffic environment, the informations such as the velocities and accelerations of multiple vehicles in front were fused, and an exponential smoothing approach was adopted to build a car-following model of connected and automated vehicles (CAVs) based on the backward-looking effect. Then the effects of number of vehicles in front and behind and the completeness of state information on the model stability were studied. The linear stability analysis was carried out and the optimal parameters of the model were determined by combining the Lyapunov's first method and linear harmonic perturbation method. Combined with the characteristics of mixed traffic environments, the car-following strategies of CAVs in different positions and states were proposed in the condition of communication information loss. The numerical simulations were carried out in three typical scenarios with different CAV penetration rates, including vehicle starting, vehicle braking to stop, and circular road. Research results show that in the scenario of vehicle braking to stop, the maximum stopping wave speed of all vehicles increases by 26.1%. In the scenario of vehicle starting, the maximum starting wave speed increases by 15.5%, and the accelerations and speeds of vehicles change more smoothly. In the circular road scenario, when the CAV penetration rate in the mixed traffic flow increases from 40% to 100%, the fluctuation time of average speed of vehicles in the larger disturbance scenario decreases by 44.8%, the wave peak decreases by 5.7%, and the wave trough increases by 19.4%, compared to the low CAV penetration rate scenarios. However, the proposed optimization strategy does not significantly improve the mixed traffic flow at a lower CAV penetration rate. Thus, in the current situation where it is difficult to construct an actual mixed traffic environment and conduct CAV real vehicle tests, the model and strategies can be employed for the car-following simulation and test verification in specific scenarios to effectively guarantee the absorption of traffic flow disturbance and stable driving vehicles in the mixed traffic environment.More>

2023, 23(3): 232-247. doi: 10.19818/j.cnki.1671-1637.2023.03.018

FullText HTML (831) PDF (25318KB) (329)

A method of heterogeneous truck platoon formation for low energy consumption based on vehicle ranking

WU Na, NIE Xiao-xiong, GE Ying-en, ZHAO Xiang-mo

Abstract: Considering the power system difference between petrol vehicles and electric vehicles, and the influence of truck type on energy-saving effect in a platoon, the energy consumption characterization model of different truck platoon was proposed based on the truck physical energy consumption model and the energy-saving relationship matrix. The energy-saving model of a truck platoon was constructed to avoid the platoon operation efficiency loss. To maximize the energy-saving benefit of platoon operation, an optimization model of the heterogeneous truck platoon for low energy consumption was built with the constraints of truck quantities, platoon lengths, and platoon ranking positions, and then the platoon sizes and vehicle ranking for each platoon were optimized. Based on numerical experiments, the energy-saving performance of heterogeneous platoon was quantitatively analyzed, and the influence rules of driving conditions, energy-saving coefficient, and the proportion of vehicle type on energy consumption were analyzed. Research results show that compared with independent driving, the energy-saving rate of the heterogeneous truck platoon is 6.5% and 6.6% in the experiments with 12 and 18 vehicles, respectively. Compared with the homogeneous platoon, the energy-saving rate of the heterogeneous truck platoon increases by 4.6% and 4.8% in the experiments with 12 and 18 vehicles, respectively. The formation of heterogeneous truck platoon does not change with the driving conditions. Compared with the constant speed condition, the energy-saving rate is about 3.6% and relatively lower in the accelerated condition. In terms of vehicle ranking principles, petrol vehicles should be arranged according to the bell principle. Electric vehicles should be inserted according to the bell principle and are used to help petrol vehicles reduce energy consumption. Finally, priority should be given to the front positions of a platoon. With the rising speed, the energy-saving rate of a platoon increases with an elasticity of 0.05, which indicates that platoon technology can bring considerable benefit on the highway scenario. With the increasing proportion of electric vehicles, the energy-saving rate of the platoon rises slightly. The energy-saving coefficient exerts the greatest influence on the energy-saving rate of the platoon, with an elasticity of 1.2.More>

2023, 23(3): 248-258. doi: 10.19818/j.cnki.1671-1637.2023.03.019

FullText HTML (299) PDF (3990KB) (99)

User Center

Journal Info

Collection of Source JournalsMore>

2023 Vol. 23, No. 3

Search Type

Search Journal

LinksMore>