交通运输工程学报

Review on ultra-high strength concrete filled steel tubes

CHEN Bao-chun, LI Li, LUO Xia, WEI Jian-gang, LAI Xiu-ying, LIU Jun-ping, DING Qing-jun, LI Cong

Abstract: To understand the research status of the ultra-high strength concrete filled steel tube(UCFST), the strength development processes in steel tubes and core concrete in concrete filled steel tube(CFST) were analyzed. A set of concise CFST classification and abbreviation methods were put forward according to the combination of different strength grades of these two materials. The basic mechanical, shrinkage and interfacial bond properties of UCFST, and the main influencing factors were summarized. The technical requirements of core ultra-high strength concrete(UHSC) were discussed, and future research directions were proposed. Analysis result shows that UCFST can be fabricated and researched in two ways: UHSC and ultra-high strength steel(UHSS). The former is the main material used in China, while the latter lags behind in its practical application is less. Although the basic mechanical properties of the UCFST have been tested, the research is incomplete and research at the structural level is rarely conducted. Research on UCFST mainly focuses on the component level, but the amount of testing is relatively small, and analysis focuses on axially loaded stub columns. There is no research on UCFST components under shear, torsion, and residual composite forces. In terms of material combinations, research attention is mostly placed on UCFST within UHSC as core concrete, followed by UCFST within both UHSC and UHSS, with only a few studies addressing other combinations. The researche on strength matching between steel tubes and core concrete has just begun, and additional studies should be proposed, focusing on the UCFST with reasonable matching. Debonding occurs between the core UHSC and steel tube because of the high autogenous shrinkage of the former. Hence, additional experimental research on the shrinkage properties and normal interfacial bond strength of the UCFST should be conducted to understand the true interfacial state. The working environment, construction conditions, and influence on the UCFST composite performance of core UHSC materials should be considered. The core UHSC materials are mainly required for ultra-high strength, low shrinkage(or micro expansion), and high fluidity, but the durability not has to be emphasized. The UHSC mix may have fibers of low volume content or even no fibers.More>

2020, 20(5): 1-21. doi: 10.19818/j.cnki.1671-1637.2020.05.001

FullText HTML (786) PDF (1266KB) (454)

Research progress on engineering properties of coarse-grained saline soil

YANG Xiao-hua, ZHANG Sha-sha, LIU Wei, YU Ze-long

Abstract: Based on the distribution of saline soil, the environmental protection feature and the importance of rational utilization of coarse-grained saline soil were expounded. In view of the single-factor salt expansion test results, combined with the multi-factor interaction salt expansion test, the effects of five factors such as soil, water, salt, temperature and force on the salt expansion of coarse-grained saline soil were analyzed. According to the comparison of on-site and laboratory collapse test results, the typical laws of collapse characteristics of coarse-grained saline soil were concluded. The results of the freeze-thaw cycle test of coarse-grained saline soil under different working conditions were also summarized, and the existing improvement methods of saline soil were refined. The research directions of engineering characteristics of coarse-grained saline soil were proposed. Research result shows that the physical and mechanical properties of coarse-grained saline soil are related to its particle composition and salt content. The granule gradation of coarse-grained saline soil is also different in different regions. Temperature, salt content, salt type, moisture content, initial density and overburden load are all key factors that affect the salt expansion of coarse-grained saline soil, while certain interactions occur among these factors. It is suggested that when coarse-grained saline soil is used as a subgrade filler, the influence of various factors on salt expansion should be carefully considered, and the factors that inhibit salt expansion should be effectively utilized. Both on-site and laboratory collapse tests are identified as effective methods to determine the saline soil collapse rate. However, the laboratory collapse tests need to be improved and standardized with consideration of the particle effect of coarse-grained saline soils. The freeze-thaw cycle test can approximately simulate the actual working conditions and directly reflect the laws of temperature, moisture and deformation of coarse-grained saline soil. In the design of test scheme, various environmental factors in the actual project should be comprehensively considered, and the combination of different factors has a significant impact on the freeze-thaw cycle test results. When improving the engineering properties of coarse-grained saline soil, the on-site materials should be fully utilized, and volcanic ash and fly ash have better improvement effects. Establishing a four-field fully coupled model of water-heat-salt-force is considered as a fundamental step to improve the salt expansion mechanism of coarse-grained saline soil.More>

2020, 20(5): 22-40. doi: 10.19818/j.cnki.1671-1637.2020.05.002

FullText HTML (561) PDF (2492KB) (2473)

Review on inspection technology of railway operation tunnels

WANG Shi-lei, GAO Yan, QI Fa-lin, KE Zai-tian, LI Hong-yan, LEI Yang, PENG Zhan

Abstract: To understand the research and application status of inspection technology for railway operation tunnels, the characteristics and inspection methods of tunnel damage were summarized. The current status of inspection technology in domestic and foreign was analyzed from five aspects: exterior state, internal state, geometric shape, high-precision ground mobile inspection robot, and data informatization. The inspection technology framework and development direction were discussed. Analysis result shows that the exterior state inspection mainly includes the camera shooting and laser scanning technology. The camera shooting system is suitable for the vehicle platform with an inspection speed of 80 km·h^-1. The laser scanning system has a compact structure and inspection speed of approximately 5 km·h^-1. The image processing and computer vision are technologies for the exterior damage recognition. Regarding the image processing technology, a potential development direction is to expand and design the damage characteristics, improve the recognition efficiency, and reduce the interference of nondamage factors. The key to the promotion of computer vision is to build an industry-level disease sample database. The ground-penetrating radar is the key technology for the internal state detection, in which the speed of ground-coupled radar is approximately 10 km·h^-1, while that of the air-coupled radar is 80 km·h^-1. The air-coupled radar detection system focuses on optimizing the antenna structure, signal enhancement, and suppression of interference from the mechanical system vibration and electrification facilities. Detection technologies such as the ground-penetrating radar, infrared thermal imaging, ultrasonic tomography, and laser defect detection are complementaries in the detection range, accuracy, and efficiency, which can constitute acomprehensive application strategy of multi-technology. The geometric shape detection mainly includes the laser scanning, laser photography, and inertial measurement technology. The measurement accuracy of laser scanning is high. Its speed is approximately 10 km·h^-1, while the laser photography speed is up to 60 km·h^-1. The system calibration and vibration compensation are crucial to improve the measurement accuracy of laser photography. In addition, based on an inertial measurement technology, the deformation detection of inverted arch uplift can be carried out. The development and promotion of a high-precision ground mobile inspection robot and the informatization of detection data are guarantee measures that adapt to the scale of tunnels and match the state accurate management.Resultssuggest that the inspection technology framework should be composed of three parts: the vehicle mounted rapid comprehensive detection, in-situ and ground mobile accurate detection, and data information platform.The development direction should focus on the rapid inspection with an air-coupled radar, fast and accurate measurement of compound deformation, high-precision ground movement inspection, intelligent damage identification, and multi-source data fusion analysis.More>

2020, 20(5): 41-57. doi: 10.19818/j.cnki.1671-1637.2020.05.003

FullText HTML (606) PDF (3127KB) (2822)

Review on connected and automated vehicles based cooperative eco-driving strategies

YANG Lan, ZHAO Xiang-mo, WU Guo-yuan, XU Zhi-gang, MATTHEW Barth, HUI Fei, HAO Peng, HAN Meng-jie, ZHAO Zhou-qiao, FANG Shan, JING Shou-cai

Abstract: To track the research progress of connected and automated vehicles(CAV) based cooperative eco-driving strategies in recent years, the influences of four factors, including the vehicle, driver behavior, traffic network and social factor on the energy consumption of CAV were analyzed. The current ecological studies on CAV were classified with vehicle, infrastructure and traveler as objects. The status-quo of 5 representative types of cooperative eco-driving scenarios were emphatically analyzed, including the eco-approach and departure at the signalized intersection, eco-cooperative adaptive cruise control, eco-cooperative driving in the ramp merging area, eco-cooperative lane changing trajectory planning and eco-routing. Analysis result shows that compared with the human driving mode, CAVs can save up to 63% fuel consumption at any traffic flow with 100% penetration rate of CAV as well as in the light traffic condition with partial penetration rate of CAV. CAVs with partial automated and connected levels can save at least 7% fuel consumption. Few existing studies consider the trajectory tracking deviation caused by the driver's response delay and automatic controller transmission delay in the case of human-machine co-driving. The existing researches assume the vehicle-to-vehicle communication(V2V) and vehicle-to-infrastructure communication(V2I) as the ideal data interaction processes. The impacts of factors such as the communication topology, transmission delay, communication failure and packet loss on the CAV based cooperative eco-driving strategies are ignored. Few existing studies discuss the eco-driving strategies in these traffic scenarios, such as the multi-lanes, shared lanes for turning and through at intersection, U-turn, as well as the mixed traffic conditions of different automated and connected level CAVs coexisting with human-driven vehicles, pedestrians and bicycles. Limited by the immaturity and imperfection of automatic driving technology and infrastructure, the test and verification work in real traffic scenarios is not carried out. The vehicle control, V2V communication, multi-vehicles collaboration, mixed traffic flow scenario, hardware-in-the-loop simulation test and real traffic scenario test will be the further development direction of CAV based cooperative eco-driving strategies.More>

2020, 20(5): 58-72. doi: 10.19818/j.cnki.1671-1637.2020.05.004

FullText HTML (892) PDF (1237KB) (1058)

Temperature deformation of CRTSⅡslab ballastless track considering interfacial initial bond defects

SU Miao, ZHU Qi-zhi, DAI Gong-lian, PENG Hui

Abstract: Aiming at the phenomenon that the temperature deformation of China railway track system(CRTS) Ⅱ slab ballastless track in Chinese high-speed railways increases with the interfacial initial bond defects, an detection experimental system for the interfacial porosity of slab ballastless track was established based on the charge coupled device(CCD) industrial camera and the computer image processing technology. The initial porosities of interface between the track slab and the cement asphalt(CA) mortar from three CRTS Ⅱ ballastless track specimens were detected. In the finite element model, the initial porosity was used to quantitatively characterize the interfacial bond states. Considering a certain amount of initial porosity at the interface, according to the test results of interface porosity and assuming these voids evenly distributing on the entire interface, the effects of interfacial initial bond defects on the temperature deformation of slab ballastless track were systematically analyzed. Research result shows that the average interfacial porosity of the three CRTS Ⅱ ballastless track slabs is 22.3%, and the bond state around the bonding interface is significantly worse than that at the central positions of track slab interface. Under the actions of positive and negative vertical temperature gradients, the CRTS Ⅱ ballastless track presents the temperature deformation modes of slab center arch and slab edges warp, respectively. Compared with the track slab without considering the interfacial initial porosity, the maximum temperature deformation of track slab increases by 7.8%-10.1% under the action of positive temperature gradient, and with the further increase of interfacial initial porosity, the maximum arch temperature deformation of track slab increases linearly. Under the action of negative temperature gradient, the increase of interfacial porosity has little effect on the temperature deformation of track slab. The interfacial initial porosity defects between the track slab and the CA mortar should be considered properly when analyzing the temperature deformation of CRTS Ⅱ slab ballastless track. The research results can provide references for analyzing the temperature deformation mechanism of CRTS Ⅱ track slab.More>

2020, 20(5): 73-81. doi: 10.19818/j.cnki.1671-1637.2020.05.005

FullText HTML (189) PDF (2220KB) (2256)

Test on impact vibration transmission and attenuation characteristics of low vibration track

CENG Zhi-ping, TIAN Chun-yu, CHEN Zhuo, WANG Jun-dong, LI Shi-ye, WU Zhi-peng

Abstract: Aiming at the problems of excessive deformation and easy failure of elastic components in the application of low vibration track(LVT) for the heavy haul railway, the bearing block was optimally designed. The short side slope was adjusted from 1∶17.00 to 1∶4.85, the backing board under the bearing block was removed, and the elastic rubber boot was integrated. The traditional LVT and improved LVT full-scale models were established to verify the design results. A wheelset with a mass of 1 120 kg was used to carry out the drop impact test at a drop height of 20 mm. The transmission and attenuation characteristics of vertical vibrations of rail, bearing block, track slab, base slab and ground along the longitudinal, vertical and lateral directions of track under the impact were compared and analyzed from the aspects of time domain and frequency domain, respectively. Research result shows that the high frequency vibration energy generated by the wheel-rail impact transmits along the longitudinal direction of track, and the low frequency energy transmits to other track structures. The vertical impact vibration attenuates continuously in the process of longitudinal and vertical transmissions, and the attenuation rate decreases gradually. In the lateral transmission process of vibration on the surface of bearing block and track slab, the transmission vibration to the outer edge increases. Compared with the traditional LVT, the whole elastic coefficient of improved LVT decreases by 21.1%, while the damping coefficient increases by 5.4%. The vibration period, attenuation time and acceleration peak of improved LVT reduce by 37.0%, 21.3% and 3.4%, respectively, and the power spectral density peak of each structure component is more than 30% smaller than that of traditional LVT. The Z vibration level at each track structure component of improved LVT is smaller than that of traditional LVT, and the Z vibration level at the ground reduces by 3.65 dB. It can attenuate the wheel-rail impact force and track structure vibration more effectively and rapidly. The vibration level is lower, and the impact effect on the environment reduces. The research result serves good references for the vibration reduction performance test verification, optimization and engineering application of LVT.More>

2020, 20(5): 82-92. doi: 10.19818/j.cnki.1671-1637.2020.05.006

FullText HTML (324) PDF (2797KB) (2247)

Influence of environmental wind on fluctuation velocity of contact wire of high-speed railway

CHEN Xiao-qiang, ZHANG Xi, WANG Ying, WANG Xin-yi, ZHONG Di-hao

Abstract: The air damping and pulsating wind aerodynamic load acting on the contact wire were respectively deduced based on the aerodynamic theory, and the aerodynamic term was added to correct the formula of fluctuation velocity of contact wire. Through the wind tunnel test and computational fluid dynamics(CFD), the aerodynamic drag coefficient under the transverse wind environment was obtained, and the variation rules of fluctuation velocity of contact wire under different air dampings were analyzed. Based on the AR model and the structural characteristics of catenary, the pulsating wind field of catenary with time and space correlation was established. The influences of pulsating wind speed and wind attack angle on the fluctuation velocity of contact wire were analyzed through the simulation. Research result shows that the air damping of contact wire caused by the static wind load is very small. When the average wind speed reaches 30 m·s^-1, the air damping acting on the contact wire is only 0.3, and the fluctuation velocity of contact wire is stable at about 549.1 km·h^-1. Therefore, the air damping will not have a great impact on the fluctuation velocity of contact wire. When the wind attack angle of incoming wind is 60°, and the average wind speed is no more than 10 m·s^-1, the standard deviation and difference between the maximum and the minimum fluctuation velocities of contact wire under the pulsating wind are less than 1 and 6 km·h^-1, respectively. In this case, the fluctuation velocity of contact wire has little change compared with the non-wind condition, and the influence of pulsating wind load on the fluctuation velocity of contact wire is not obvious. When the wind speed reaches 40 m·s^-1, the average fluctuation velocity of contact wire decreases by 39.39 km·h^-1 compared with the non-wind condition, and the standard deviation and difference between the maximum and the minimum fluctuation velocities reach 11.84 and 75.98 km·h^-1, respectively. At this point, the fluctuation velocity of contact wire decreases and oscillates sharply, and the minimum is as low as 474.16 km·h^-1. Therefore, the larger the wind speed is under the pulsating wind, the more significant the impact of pulsating wind load is on the fluctuation velocity of contact wire. When the wind speed is maintained at 30 m·s^-1, and the attack angle of incoming wind is 0°-30°, the standard deviation and difference between the maximum and the minimum fluctuation velocities are less than 1 and 5 km·h^-1, respectively. In this case, the pulsating wind load has little influence on the fluctuation velocity of contact wire. When the wind attack angle is 90°, the standard deviation and difference between the maximum and the minimum fluctuation velocities reach 12.38 and 73.19 km·h^-1, respectively. At this point, the fluctuation velocity of contact wire decreases and oscillates sharply, and the minimum is as low as 472.91 km·h^-1. Therefore, under the action of pulsating wind, the more the incoming wind tends to the horizontal direction, the less influence it has on the fluctuation velocity of contact wire.More>

2020, 20(5): 93-104. doi: 10.19818/j.cnki.1671-1637.2020.05.007

FullText HTML (253) PDF (1296KB) (249)

Effective protection distance of HDPE board fence in Golmud-Korla Railway

ZHANG Kai, WANG Qi-cai, ZHANG Xing-xin, ZHAO Pei-wen

Abstract: The field sand test section of the Golmud-Korla Railway was taken as the research object, and the sand flow field around the HDPE board fence was studied through the numerical simulation. The relationships between the effective protection distance of HDPE board fence and its porosity and height under different initial wind speeds was given. Research result shows that when the air flow passes through the HDPE board fence, the air flow speed decreases rapidly in front of the fence, and recovers quickly after the fence. After a certain distance, it gradually returns to the initial wind speed. The air flow speed is distributed in a V shape. The trends of air flow speed reduction and increase gradually decrease with the increase of the porosity of HDPE board fence. Under the same porosity, when the initial wind speed is 6, 24 m·s^-1, respectively, the difference between the recirculation zone on the leeward side of HDPE board fence is 4.5 times of the height of HDPE board fence. When the porosity is 60%, the minimum air wind speed is 8.9 m·s^-1, and the backflow on the leeward side of HDPE board fence disappears. As the porosity of HDPE board fence increases, the minimum air flow speed gradually increases. The porosity of HDPE board fence has a limit porosity that the leeward side of fence does not occur, which is 40%-60%. When the porosity is less than 50%, the effective protection distance increases gradually as the porosity of HDPE board increases. When the porosity is greater than 50%, as the porosity of HDPE board increases, the effective protection distance decreases gradually. When the porosity tends to 100%, its effective protection distance is almost 0. Therefore, the optimal porosity of HDPE board fence is 50%. As the height increases, the distance of returning to the initial wind speed after the fence increases. Under the same wind speed, the effective protection distance of HDPE board fence with 50% porosity is 1.35 times of that with 25% porosity. The HDPE board fence with 40%-50% porosity is suggested to use when installing on site. The height of fence should appropriately increase if the economy permits, so as to ensure the roadbed being free from the sand erosion.More>

2020, 20(5): 105-115. doi: 10.19818/j.cnki.1671-1637.2020.05.008

FullText HTML (265) PDF (1615KB) (282)

Improvement of fracture failure criterion of body adhesive structure

NEI Jing-xin, LOU Fei, JI Jun-dong, LI Feng

Abstract: The existing failure criterion cannot meet the real failure prediction of bonding structures. Thus, a stress-based fracture failure criterion was established by combining an experimental test and simulation analysis. Five groups of ISR-7008/aluminum alloy bonding joints with typical pull-shear ratios were designed. Quasi-static tensile tests were carried out on the five groups of bonding joints with different pull-shear ratios. The initial fracture load and the maximum fracture load were obtained. The initial location of adhesive fracture failure point was determined. A simulation model of the bonding joint was developed. The initial fracture load was applied to the simulation model. The stresses at the initial fracture point in the failure area of the five joints with typical pull-shear ratios were extracted. Through the ratio and linear combination of various stresses at the failure point, an equivalent stress calculation formula was obtained. Based on the equivalent stress calculation formula, a unified failure criterion suitable for the initial failure and subsequent failure was obtained. A verification test plan was designed. The validity of failure criterion was illustrated by comparing the test results and simulation results. Analysis result shows that the fracture load obtained by the simulation analysis in the 75°scarf joint is 1 717.6 N, while the fracture load obtained by the test is 1 936.4 N. The relative error is 11.3%. The simulation results are consistent with the failure process of bonding layer tested in the experiment. The failure criteria established in this study are verified. The unification of initial failure criterion and subsequent failure criterion of the bonded structure is realized in the stress-based failure criterion established in this study. The failure of bonding joints under complex stress conditions can be predicted more accurately. The simulation problem of the thick layer of elastic adhesive is solved by this failure criterion. A certain reference for the strength design of bonding structure in practical engineering applications is provided.More>

2020, 20(5): 116-124. doi: 10.19818/j.cnki.1671-1637.2020.05.009

FullText HTML (255) PDF (1482KB) (311)

Hunting control of high-speed train using traction motor as dynamic absorber

ZHANG Wei-hua, LUO Ren, SONG Chun-yuan, FAN Jun

Abstract: The traction motor of power bogie of CR400BF Fuxing high-speed EMU adopted the unique four-point elastic suspension mode, and the transverse oil dampers and bump stops were installed between the motor and bogie frame. For the first time, the traction motors were used as dynamic absorbers to control the hunting stability and hunting frequency of the bogie, and further to avoid the elastic modal resonance of car body. Considering the nonlinearity of suspension parameters and wheel/rail contact, a nonlinear multi-body dynamics simulation model of Fuxing EMU was established. The influence of key parameters on the hunting was analyzed by the suspension modal calculation and dynamics time-domain simulation of power vehicle. Based on the principle of using the traction motor as a dynamic absorber, the lateral stiffness of motor node and the damping of oil damper were optimized. Considering the random wheel/rail matching factors in the EMU operation, 400 wheel/rail random matching states were combined to analyze the dynamics performances of the EMU. The long-term dynamics tracking test of the EMU on line was carried out, and the development of hunting phenomenon of power bogie was obtained. The simulation and test results show that the frequency spectrum of bogie frame's lateral acceleration under the motor's elastic suspension changes from a single peak with the hunting frequency as the main frequency to double peaks with the main frequency on both sides of the hunting frequency, which indicates that the motor acts as a dynamic absorber. Taking the motor as a mass damper can improve the hunting stability of motor vehicle. The nonlinear critical speeds under typical wheel/rail matchings with different equivalent conicities exceed 500 km·h^-1. The highest hunting frequency of motor vehicle is around 6 Hz, which is far away from the diagonal distortion mode frequency 8.5 Hz of car body. Therefore, the elastic resonance of car body caused by the bogie hunting is avoided. Under the excitation of random track irregularity, the transverse acceleration at the end of bogie frame is less than 0.5g, the spelling index is less than 2.5, and the wheelset lateral force and derailment coefficient meet the requirements.More>

2020, 20(5): 125-134. doi: 10.19818/j.cnki.1671-1637.2020.05.010

FullText HTML (302) PDF (1509KB) (263)

High-speed train gearbox housing vibration test based on small roller high-frequency excitation

ZHU Hai-yan, WANG Chao-wen, WU Ping-bo, CENG Jing, XIAO Qian

Abstract: To analyze the vibration characteristics and fatigue damage of a high-speed train gearbox housing, a small roller high-frequency excitation test stand was applied to study the vibration characteristics of gearbox housing under different vertical loads and speeds. In the test, the surface of roller was processed to the 13 th-order polygon with a radial jump amplitude of 0.05 mm, which could be equaled to a 20 th-order actual wheel polygon. The stress accumulation damage of gearbox housing per unit time was analyzed by the rain flow counting method and Miner linear damage rule. Analysis result indicates that effected by the resonance of the gearbox housing, the root mean square(RMS) values of transverse and vertical accelerations of each measuring point of the gearbox housing are minimal under different vertical loads and speeds at a running speed of the high-speed train of 200 km·h^-1. The RMS values of transverse and vertical accelerations of most measurement points are maximal when the vertical load is 23 t. The resonance of gearbox at the natural frequency of 573 Hz can be stimulated by the resonance frequency of test bench at a test speed of 100 km·h^-1 and main excitation frequency of the 20 th-order wheel polygon frequency shift of approximately 580 Hz at a test speed of 300 km·h^-1. The cumulative stress damage per unit time at each measuring point of gearbox housing largely fluctuates when the initial speed of wheel is between the speed grades of acceleration from 0 to 200 km·h^-1 and deceleration from 300 to 0 km·h^-1. The stress cumulative damage per unit time of each measuring point in the other speed grades has small fluctuations. The maximum value of stress damage per unit time is obtained in the tooth surface observation hole of a large gearbox housing, 3.72×10^-10, while the minimum damage is located above the bearing of a small gearbox housing, only 8.29×10^-18. Thus, the vibration acceleration of gearbox housing is largely affected by the resonance of gearbox and the deceleration operation and speed grades of test stand. However, the vertical load has a small influence on the stress cumulative damage per unit time under the conditions of non-resonance of gearbox housing, without deceleration operation, and the same speed grade of test stand.More>

2020, 20(5): 135-150. doi: 10.19818/j.cnki.1671-1637.2020.05.011

FullText HTML (322) PDF (5024KB) (2455)

Multi-scale convolution intra-class transfer learning for train bearing fault diagnosis

SHEN Zhang-qing, WANG Xu, WANG Dong, QUE Hong-bo, SHI Juan-juan, ZHU Zhong-kui

Abstract: Considering the inconsistent distribution of bearing vibration data collected under different working conditions, the generalization ability of traditional deep learning model decreases. A multi-scale convolution intra-class adaptive deep transfer learning model was proposed. The spectrum of vibration data was analyzed using the modified ResNet-50. The middle-level features were obtained. A multi-scale feature extractor was developed, the middle-level features were processed, and the high-level features were generated. The high-level features were used as the inputs of classifier. The pseudo-labels were computed, and then the conditional distribution distances of vibration data collected under variable working conditions reduced for the intra-class adaptation. To verify the generality and superiority of model, the proposed method was employed to analyze a train wheelset bearing dataset and the Case Western Reserve University dataset under variable working conditions. Analysis result indicates that the high-level features of samples with the same label in different domains are properly aligned. More satisfactory fault diagnosis accuracies are obtained by the proposed model. In six fault diagnosis cases of train bearing under variable working conditions, the average diagnosis accuracy of the proposed model is 90.75%, approximately 10% higher than those of traditional deep learning models, while the recall rate is 0.927. In twelve fault diagnosis cases of Case Western Reserve University dataset under variable working conditions, the average accuracy obtained by the proposed model is 99.97%, approximately 10% higher than those of traditional models. The conditional distribution discrepancy between different domains reduces by using the pseudo-labels. The inconsistency problem of data distribution of source domain and target domain is properly addressed. The high-level features of samples from different scales can be aligned by the multi-scale feature learner. The generalization and robustness of the model largely increase. In conclusion, the proposed model has a high potential for the train bearing fault diagnosis under variable working conditions.More>

2020, 20(5): 151-164. doi: 10.19818/j.cnki.1671-1637.2020.05.012

FullText HTML (287) PDF (2784KB) (2366)

Quasi-static force analysis and longitudinal load distribution law of heavy haul freight coupler

ZHU Tao, YIN Min-xuan, XU Jing-tao, WANG Chao, XIAO Shou-ne, YANG Guang-wu

Abstract: To evaluate the change in the force state of a heavy haul freight coupler under the combined action of coupler clearance, gravity, and longitudinal traction force, a quasi-static force analysis for the coupler was carried out and the longitudinal load distribution law was studied. A strain test at the roots of the upper and lower traction flanges of the knuckle was designed. The relationship between the elastic stain of measuring points and traction force was obtained. The change trends of bearing degrees of the upper and lower traction flanges were analyzed. A detailed force analysis of coupler was carried out. An analytical solution of equivalent forces on the load transfer sites was derived. The boundary conditions of the simulation under different traction forces of coupler were obtained. A simulation analysis of coupler was carried out. The response curve of node strain with the traction force was obtained. Through comparison to the results of strain test, the reliabilities of analytical solution of the equivalent forces on the load transfer sites and simulation model were demonstrated. The influences of some key parameters on the relationship between the traction force and equivalent force were analyzed. The equivalent forces refer to the forces acting on the inner wrist surface and upper and lower traction flanges. Research result shows that the upper shocking platform is stressed when the traction force is smaller than 13.5 kN. The upper and lower traction flanges of the coupler are loaded simultaneously when the traction force is 13.5-1 725.0 kN. With the gradual increase in the traction force, the bearing ratio of lower traction flange gradually decreases and approaches 0.53, while the bearing ratio of upper traction flange increases and approaches 0.47. The bearing ratio is linear with respect to the system parameters. The acting position of equivalent force on the inner wrist surface of knuckle has a large influence on it. These results can be used as a basis for studying the fatigue crack initiation and propagation simulations of couplers and have a high guiding significance for the couplers' service safety.More>

2020, 20(5): 165-175. doi: 10.19818/j.cnki.1671-1637.2020.05.013

FullText HTML (388) PDF (1282KB) (285)

Multi-mode passenger flow sharing characteristics of highway-rail composite intercity corridor

XU Ming-fei, LI Chang-ling, WANG Yuan-qing, ZHOU Wei

Abstract: Four node cities, including Xianyang, Weinan, Huangling and Yan'an within 350 km from the central city of Xi'an in Shaanxi Province were selected. The passenger flow characteristic parameters such as travel volume, time inside the vehicle, ticket price or toll of high-speed railway, ordinary railway, long-distance bus and car on the highway were collected. Various methods of intercity passenger flow sharing analysis were summarized. A distance transfer curve model and multivariate Logit model were constructed and were calibrated by curve fitting, trial calculation and regression analysis. According to the model calibration results, the sensitivities of passenger flow sharing rate to distance, time and cost were analyzed, respectively. The sharing characteristics of regional intercity multi-mode passenger flow were obtained, and the relevant suggestions on the planning and management of intercity corridors were given. Analysis result indicates that the fitting results of sharing rate-distance transfer curves of three modes, including high-speed railway, ordinary railway and car on the highway, are ideal. The determination coefficients are all above 0.94. When MNL model takes 90-150 min out of the vehicle, the fitting effect is preferable and the determination coefficients are all above 0.79. The determination coefficients reach the peak values when the time value is 50-70 yuan·h^-1. With the increase of intercity travel distance, the travelers choose to transfer from car on the highway to intercity railway travel. A high-speed railway has more advantage than an ordinary railway. The sharing rate of the car on the highway between Xi'an and Xianyang in a near distance is 96.91%. The sharing rate of the high-speed railway between Xi'an and Yan'an in a long distance is 53.66%, and that of the ordinary railway is 30.58%. Taking the outside time of 120 min as an example: the ranges of impedance coefficient of high-speed railway, ordinary railway, long-distance bus and car on the highway are 0.029-0.044, 0.034-0.042, 0.030-0.040 and 0.028-0.048, respectively, car and high-speed railway have a larger growth range, making both services more sensitive to travel costs. The impedance coefficients of the four modes are 0.038-0.042 under the condition that the time value is 60 yuan·h^-1. There is no significant difference between the four services on the generalized time senisitivity of travel. It is suggested that more passenger flow sharing rules of inter city corridors in urban agglomerations should be explored. The details of an urban end of an inter city travel chain should be considered accurately to guide the macro planning and management of intercity corridors preferably.More>

2020, 20(5): 176-186. doi: 10.19818/j.cnki.1671-1637.2020.05.014

FullText HTML (310) PDF (592KB) (267)

Vehicle scheduling model considering individual driving speed deviation

ZHANG Na, YANG Qi, HU Fei-hu, JIANG Xin-yu, LIU Yong-xiong

Abstract: In view of driving speed deviation, a vehicle scheduling model was established for multi-drivers, multi-vehicles, multi-materials, multi-depots, and multi-demands targeting at the shortest overall transportation time, the lowest overall transportation cost and the least multi-objective overall transportation time and cost, respectively. The effects of individual driving speed deviation on the above targets were studied. The driver parameters were input into the gene coding of the genetic algorithm. The constraints of driver uniqueness, initial location, and the supply and demand quantities of materials were established to ensure that the distribution scheme of drivers in each gene was feasible and the material transportation did not exceed the total supply and demand. A genetic algorithm was applied to solve the vehicle scheduling schemes for each target with and without driving speed deviation under the condition of randomly assigned drivers. Calculation result shows that the optimized scheduling schemes satisfy all the constraints in the model. For the three optimal schemes, the driver assignments are different, indicating that the target function is affected by the driving speed deviation. The dispatching results of each target with driving speed deviation are superior to those without driving speed deviation. The difference ratios of the three objective functions are 3.5%, 2.96% and 1.13%, respectively, which shows that the driving speed deviation has a certain influence on the solving quality. The target scheduling results of the driver's random assignment are inferior to the corresponding optimal results. The different ratios of the three objective functions are 3.91%, 2.47% and 1.98%, respectively, showing that the dispatching efficiency is affected by the driving speed deviation, and optimizing the driver allocation scheme can reduce the overall transport time and cost. Analysis result shows that the vehicle scheduling scheme, which is more in line with the scheduling target, closer to reality, and more economical and time-saving, can be obtained by allocating drivers reasonably according to the specific dispatching target.More>

2020, 20(5): 187-197. doi: 10.19818/j.cnki.1671-1637.2020.05.015

FullText HTML (277) PDF (607KB) (296)

Evaluation method of security risk on crowded metro station

WANG Yi-min, LUO Yue, YU Heng, CHEN Jia-cheng, HUANG Xin

Abstract: In order to accurately evaluate the security situation of crowded metro station, the security risk factors of crowded metro station were identified and analyzed on the basis of considering the subjective and objective hazard sources through literature search, case collection and questionnaire. A 4-level security evaluation system with 55 indicators was established. The risk types included fire, stampede, terrorist attack, flood, earthquake and so on. The calculating method of indicator weights at all levels, the acquisition method of indicators scores, the types of security grade and the indicators score ranges were clarified. Based on the extension theory, the security score calculation method of four-level risk indicators and the security risk grade judging criteria and evaluation method of crowded metro station were established. The security evaluation method was programmed and calculated by using MATLAB. The Tiyuxilu Station of Guangzhou Metro Line 3 was evaluated rapidly. Calculation result shows that the correlation matrix between the first-level security indicator and each security level of Tiyuxilu Station is(-0.057 6, -0.462 4, -0.588 2, -0.628 1)^T, and the correlation matrix between the stampede accident of second-level indicator and each security level is(-0.354 8, -0.741 5, -0.724 5, -0.690 4, -186 5)^T. According to the principle of maximum correlation degree, Tiyuxilu Station is generally in a safe state, but there are potential security hazards in congestion/stampede accident, evacuation environment, and management countermeasures, so the good operational security risk control measures of the station can reduce the impact of unsafe factors. The evaluation method is conducive to accurately understand the security situation of metro station, timely discover the potential security hazards, and formulate the countermeasures to ensure the safe operation of metro station.More>

2020, 20(5): 198-207. doi: 10.19818/j.cnki.1671-1637.2020.05.016

FullText HTML (608) PDF (782KB) (308)

Characteristics and vulnerability of rail transit network besed on perspective of passenger flow characteristics

MA Chao-qun, ZHANG Shuang, CHEN Quan, CAO Rui, REN Lu

Abstract: In order to improve the objectivity of vulnerability assessment of urban rail transit network, passenger demand characteristics were integrated into the calculation of network vulnerability. Based on the static topological structure of urban rail transit network established by using the Space L method, the weighted network of rail transit was established with passenger flow as the weight. Based on the passenger flow index, the station connection strength and weighted node betweenness were proposed to reflect the dynamic network structure characteristics and measure the interaction strength between nodes. Aiming at the spatial-temporal characteristics of passenger flow in urban rail transit network, combined with the demand characteristics of network passenger flow, the passenger effective path subgraph and OD loss rate of network passenger flow under the condition of station failure were defined by using the maximum travel consumption tolerance threshold to evaluate the vulnerability of urban rail transit network. Taking Xi'an urban rail transit network as an example, the features and vulnerability of urban rail transit network were interpreted from the perspective of passenger flow characteristics. Research result shows that the current rail transit network in Xi'an has the characteristic of small world network and its average path length is 10.7. Xiaozhai Station and Beidajie Station are the key nodes of the network, their connection strengths are 166 795 and 149 059, respectively, and their weighted node betweennesses are 0.365 and 0.369, respectively. The interruption of the two stations will result in 40.1% and 39.4% reduction in network efficiency. The passenger travel tolerance threshold greatly affects the importance ranking of the stations in the network. With the increase of passenger travel tolerance threshold, the network vulnerability gradually decreases. The correlation between vulnerability and betweenness is stronger than those with degree and intensity. With the increase of travel tolerance threshold, the correlation between weighted betweenness and vulnerability gradually decreases. Therefore, the calculation indicators and methods proposed in this paper highlight the impact of passenger flow characteristics and passenger demand on the vulnerability of rail transit network, which can well reflect the functional characteristics of rail transit network.More>

2020, 20(5): 208-216. doi: 10.19818/j.cnki.1671-1637.2020.05.017

FullText HTML (317) PDF (609KB) (329)

Airline flight frequency optimization based on multiple travel paths

LE Mei-long, ZHENG Wen-juan, HU Yu-ming

Abstract: A multiple-layer network was abstracted from the airline's air transport network, and a two-stage planning model was built to determine the flight frequency along a certain route by an airline for a specific city pair. In the first stage, a negative utility function of travel was constructed on the basis of the passengers' selection behavior by considering their perception of travel time, transfer time, delay time, and ticket price. Subsequently, a polynomial Logit model was adopted to create a route selection model in order to calculate the probability of passengers selecting a certain route by an airline for a specific city pair. In the second stage, a linear planning model was established to determine the flight frequency from the airline's perspective. The overall objective was to maximize the total revenue, the multiple travel paths, the total carrier capacity of the airline, and the balance between the carrier supply and demand for each path were considered. An iterative algorithm was presented to solve the proposed two-stage model. Analysis result shows that the convergence can be achieved after 8 iterations, and thus, the optimal solution can be reached within a short time. As the solutions converge, the proposed two-stage planning model prioritizes the routes with the highest revenue to improve the overall revenue in cases where there is market competition and insufficient overall capacity. For the routes with multiple segments, the two-stage model can more clearly present the role of the passengers' selection behavior related to the flight frequency determination. For the routes with only one segment, there is less variation in demand with respect to the change in the flight frequency. As the number of iterations increases, the demand tends to become decreasingly sensitive to the flight frequency. For the routes with multiple segments, the variation in the demand with change in the flight frequency is considerably higher in the cases when the total demand is fixed. Conversely, the demand decreases sharply when the flight frequency remains unchanged due to the market competition. Therefore, the presented model and algorithm can effectively improve the airline revenue.More>

2020, 20(5): 217-226. doi: 10.19818/j.cnki.1671-1637.2020.05.018

FullText HTML (235) PDF (557KB) (236)

UKF-based three-dimensional track generation method for digital track map

TAO Wei-jie, CAI Bo-gen, LIU Jiang, WANG Jian, SHANGGUAN Wei

Abstract: To meet the requirements of digital track maps for the satellite-navigation-system-based train positioning, an unscented Kalman filter(UKF)-based track estimation method was proposed and a three-dimensional digital track map for railway tracks was generated. For the three basic curve elements(straight line, transition curve, and circular arc) in the horizontal profile of railway track, a mileage-parameterized Fresnel integral model was used for a unified modeling. For the straight line and curve in the vertical profile, a quadratic curve model was used for modeling. The states(mileage, three-dimensional coordinates) and parameters(heading, curvature, curvature rate, slope, and slope rate) of models were jointly estimated using the UKF. The normalized innovation squared and estimated distance error were introduced as the criteria to segment the track. The three-dimensional railway track was generated using the breakpoints along with the geometric parameters. The discrete point, cubic polynomial, and proposed Fresnel integral methods were compared by using the simulated horizontal track data. The Fresnel method was verified by using the 14.7 km field data from the Qinghai-Tibet Railway Line. Simulation result shows that the mean horizontal distance errors are below 0.024 m for all three methods under the same error requirement. However, the Fresnel method uses the fewest break points, with a data reduction rate of 99.76%. In addition, the maximum chainage error of Fresnel method is the smallest, which decreases from 0.964 m to only 0.060 m, with a reduction of 93.77%. The heading and curvature of Fresnel method are considerably more accurate than those of the cubic polynomial method, which are closer to the true value. The field data test results demonstrate that the Fresnel method can use 22 and 20 break points with their parameters to generate the horizontal and vertical curves, respectively. The mean distance errors of horizontal and vertical curves are below 0.03 m, while the maximum accumulative mileage error is only 0.078 m, which indicates high accuracies of both position and geometry.More>

2020, 20(5): 227-236. doi: 10.19818/j.cnki.1671-1637.2020.05.019

FullText HTML (116) PDF (731KB) (247)

User Center

Journal Info

Collection of Source JournalsMore>

2020 Vol. 20, No. 5

Search Type

Search Journal

LinksMore>