交通运输工程学报

Bearing capacity safety of hollow slab bridge under heavy traffic load

HUANG Ping-ming, YUAN Yang-guang, ZHAO Jian-feng, HAN Wan-shui, LI Yong-qing, WU Juan

Abstract: Based on the 18-month weigh-in-motion (WIM) data of Xuanda Expressway in Hebei Province, the load data of extra-heavy trucks were extracted, the key load parameters including the mass, speed, arrival time and lane distribution of vehicle were investigated, the types of extra-heavy trucks were classified, and the axle load distributions were analyzed. The load cases of 883 extra-heavy trucks were dynamically and visually simulated by using Bridge Dynamics Analysis System, the responses of hollow slab bridges were compared with the design vehicle load effects, and the differences between extra-heavy truck loads and corresponding design loads were investigated. The bearing capacity safety of hollow slab bridges under heavy traffic was investigated by considering the effect combination of dead load and extra-heavy truck load. Analysis result shows that the ratios of extreme values of positive bending moment and shearforce to the corresponding design values are 2.09 and 1.97, respectively, which indicates that the maximum extra-heavy truck loads are obviously higher than vehicle design loads.the ratios of average values of positive bending moment and shear force to the corresponding design values are close to 1.0, which indicates that the average extra-heavy truck load is close to the design vehicle load. The average values of evaluation indexes of bending and shearing bearing capacities fluctuate around 0.50 and 0.40, respectively, the extreme values fluctuate around 0.72 and 0.50, which indicates that the hollow slab bridge at the current design level can satisfy the operation safety requirement under heavy traffic, and the bending capacity has a higher redundancy than the shearing capacity. The evaluation indexes of bearing capacity don't change obviously with the increase of span length, which indicates that the redundancy of bearing capacity remain stable when the span length increases.More>

2017, 17(3): 1-12.

FullText HTML (144) PDF (2382KB) (1604)

Vehicle loading spectrum and fatigue truck models of heavy cargo highway

ZHU Zhi-wen, HUANG Yan, XIANG Ze

Abstract: In order to research vehicle loading spectrum and fatigue truck models of heavy cargo highway, the multi-period traffic flow data based on a weight-in-motion system located on Pingsheng Bridge in Foshan were used to present 10 representative vehicle types. The wheelbases, masses, axle loads and overload data of representative vehicle types were analyzed, the distributions of vehicle types and axle loads on each lane were studied, and vehicle loading spectrum was proposed to evaluate the fatigue performance of steel bridge. The six-axle truck with the largest fatigue loading rate was taken as prototype, the fatigue truck model and simplified fatigue truck model of unidirectional heavy load lane of bridge were proposed based onthe equivalent rule of fatigue damage. Calculation result shows that Pingsheng Bridge presents the typical features of heavy cargo highway, the average daily traffic is 45 065 veh and 2.3 times as large as 20 000 veh in AASHTO LRFD. The proportion of fatigue vehicles is 51.6% in all traffic flow and 2.6 times as large as 20.0% in AASHTO LRFD. The proportion of trucks is 45.2% in all fatigue vehicles, they mainly distribute on heavy lanes, overload trucks accounts for 30%-70% of corresponding vehicle types, and the maximum mass of overload truck reaches 132.5 t. The overload rate of two-axle trucks is 29.0%, the equivalent mass is 17.5 t, and the equivalent mass of rear axle is 12.1 t, therefore, the fatigue loading effect of two-axle trucks should not be ignored. Compared with AASHTO LRFD's five-axle standard fatigue vehicle model (the front axle load is 2.6 t, and the single axle loads of mid and rear two-axle group are 5.4 t) and the simplified fatigue vehicle model (the front axle load is 2.6 t, and the mid axle load and rear axle load are 10.8 t), the total mass of six-axle unidirectional fatigue truck model proposed in this paper is 33.1 t, the front axle load is 3.6 t, and the single axle loads of mid two-axle group and rear three-axle group are 5.9 t. The front axle load of simplified one-direction fatigue truck model is 3.6 t, the mid axle load is 11.8 t, and the rear axle load is 17.7 t. The total mass of six-axle fatigue truck model proposed for the heavy lanes reaches 36.5 t, the front axle load is 4.0 t, and each axle load of conjoint-axle is 6.5 t. The front axle load of simplified fatigue vehicle model for the heavy lanes is 4.0 t, the mid load is 13.0 t, and the rear axle load is 19.5 t.More>

2017, 17(3): 13-24.

FullText HTML (165) PDF (1919KB) (1090)

Linear-elastic analysis method of ultimate bearing capacity of dumbbell-shaped CFST arch rib

YANG Lu-feng, JIE Wei-wei, ZHENG Jian, ZHANG Wei

Abstract: In order to improve the computational efficiency of ultimate bearing capacity of dumbbell-shaped CFST (concrete filled steel tube) arch rib, a high-efficiency self-adaptive elastic modulus reduction method (EMRM) was proposed to analyze the ultimate bearing capacity. Based on the continuity conditions and the plastic bearing property of section, the correlation equations of compressing-bending capacity of dumbbell-shaped component for CFST were established, and the corresponding homogeneous generalized yield function (HGYF) wasdetermined by means of regression analysis. A linear-elastic finite element iterative model of arch rib was developed by using simplex beam element with combined material parameters, and the elastic modulus of highly loaded element was reduced through self-adaption to simulate the structural stiffness damage in the loading process, so as to confirm the ultimate bearing capacity of arch rib. The proposed method was compared with model test, non-linear finite element method and equivalent beam-column method. Calculation result shows that the calculation result of HGYF is stable and reliable, and the impact of initial loads on the calculation result of traditional generalized yield function is overcomed. The proposed method has higher accuracy and efficiency than the nonlinear finite element method, the stable ultimate bearing capacity is obtained by only small amount of discretized meshes and iteration steps, the relative error is less than 3% compared with test result data, and the computation time is less than 16 s. Compared with the circular section arch rib, the dumbbell-shaped CFST arch rib has better bearing property, and the main influence factors are rise-span ratio, steel ratio and loading condition. The increasing speed of ultimate bearing capacity reduces with the increase of rise-span ratio. With the increase of steel ratio, the ultimate bearing capacity increases almost linearly. The larger the ratio of concentrated load to uniform load is, the less its influence on the bearing capacity is. The axial force and bending moment are the governing internal forces of arch rib, while the bending moment becomes more significant with the increase of rise-span ratio.More>

2017, 17(3): 25-35.

FullText HTML (123) PDF (651KB) (849)

Water-salt migration laws of aeolian sand subgrade in desert area

HU Jian-rong, ZHANG Hong, ZHANG Hai-long, YAN Xiao-hui, LI Liang

Abstract: Based on aeolian sand subgrade soil in desert area, the chemical compositions of samples at different depths at the typical pavement-distressed section were tested, and the migration characteristics of salinity and moisture in subgrade were analyzed based on soil water potential principle.The soil was layered filling with the compaction degree of 95%, the moisture content and electric conductivity of soil pillar model were tested by using self-made experimental device, and the influence of temperature gradient on the water-salt migration rules in subgrade under the condition of optimum water content was analyzed.Research result shows that the subgrade soil of distressed section is fine sand, and sodium sulfate and sylvite are main salt components in both base and subgrade soil.Na₂SO₄ ·10H₂O generates at low temperature -6 ℃-0 ℃to lead to the volume expansion of soil.The salt expansion of subgrade soil and the upheaval destruction of pavement are aggravated because of the entrance of external moisture.After a week of subgrade compaction, the water content reduces by 1%-2% and the content of sulfate radical reduces by 0.05%-0.06% at depth of 5 cm, and the water content improves by 0.5%-0.8% and the content of sulfate radical reduces by 0.12%-0.14% at depth of 35 cm.Under the dual-action of gravity potential and compaction, quick and obvious water-salt stratification occurs in the uniform soil.Under the external temperature, the temperature difference at the depth of 25 cm is 20℃-30℃, but the difference is about 1℃ when the depth is more than 25 cm.So the temperature gradient variation gradually decreases with the increase of depth, and becomes zero eventually.The distributions of water and salt first decrease and then increase with the depth in aeolian sand subgrade, and appear hooklike rule.The water-salt migration in aeolian sand subgrade results from the mixing effect of gas-liquid states.The hydrosphere migration is primary within the depth of 10 cm from subgrade top surface under high temperature.However, because the effective access channels of capillary water can not form below 10 cm in aeolian sand composed of fine sand, the moisture mainly migrates in the form of pellicular water.The liquid water with salinity rises under the cooling effect to lead to salt accumulation effect on the top surface of subgrade.More>

2017, 17(3): 36-45.

FullText HTML (193) PDF (3726KB) (483)

Dynamic response of anchors-supported slope under earthquake

HAO Jian-bin, GUO Jin-yang, ZHANG Zhen-bei, LI Jin-he

Abstract: In order to investigate the dynamic response characteristic and failure mode of anchor in a soil-filled slope, a shaking table mode experiment on clathrate anchors-supported soil slopes was carried out. Sinusoidal motions were used as the incident waves, and the acceleration of slope and the axial strain of anchor were monitored. Analysis result shows that under the same vibration frequency, the axial strain amplitude of anchor increases with the increase of peak acceleration. When the peak acceleration is low, the slope is stable, and the strains of anchors recirculate between a positive value and a negative value. The maximum and minimum strains of anchor are unstable and change slightly with the increase of the peak acceleration, but the slope is still stable. When the peak acceleration reaches to the rupture limit, the axial strains of anchors are no longer regular, and the axial strains of key points on the sliding surface greatly and suddenly change, and the relative displacement is very obvious between the sliding body and the stable body. When the peak acceleration is less, the strains of middle and lower anchors are larger, and the strains of middle anchors are about two times as much as the value of top anchor. When the peak acceleration increases, the strains of top anchors gradually increase, and the middle andupper anchors bear loads mainly. When the peak acceleration reaches to the rupture limit, the maximum dynamic strains of anchors increase sharply, and the changing amplitudes of strains of middle anchors are greatest, a clear gap shows between the sliding body and the sliding bed, and the anchors are pulled out. Obviously, the traditional design idea"strengthening slope waist and reinforcing slope toe"is suitable for the areas of low seismic fortification intensity, and the top and middle anchors should be lengthened properly in the areas of high seismic fortification intensity.More>

2017, 17(3): 46-55.

FullText HTML (141) PDF (942KB) (614)

Comparison of measuring accuracies of tunnel displacements with RDM method and 3D measurement method based on total station

LUO Yan-bin, CHEN Jian-xun, ZHAI Yu-hui

Abstract: Based on geodesy principles and error propagation laws, the remote distance measurement (RDM) method and 3Dcoordinate measurement method of based on free stationing of total station were analyzed, the accuracy analysis models of tunnel displacement based on the two measuring methods were established, and the mean square error was used to evaluate the measuring accuracy. The formulas of mean square error on the two methods were deduced. Athree-lane highway tunnel was taken as an example to compare and verify the measuring accuracies of the two methods. In RDM method, through gaining the horizontal distance, the elevation difference and slant distance of two random measurement points were gained by the triangulate height measurement principle and the cosine theorem, and the tunnel displacement was obtained according to the trigonometric and geometric relationship between two random measurement points. In 3D measurement method, the directions and distances of several known points were observed from the random observation point, the coordinates of random measurement points were calculated by using the coordinate conversion, and the tunnel displacement was obtained according to the coordinates of random measurement points. Analysis result shows that the accuracy evaluating formulas of RDM method and 3D measurement method computing tunnel vault settlement are same, however, the formulas computing tunnel horizontal convergence are different, and the accuracy of RDM method is higher than 3D measurement method. When the distance of total station and measured profile increases, the measuring accuracy difference of two methods increases. When the distance is 100 m, the accuracy difference increases to 0.43 mm. In three-lane highway tunnel, when the distance is 40-60 m, the measuring accuracies of tunnel horizontal convergences for the two methods are highest, the accuracy of RDM method can reach 0.61-0.68 mm, and the accuracy of 3D measurement method can reach 0.78-0.84 mm. The curves of tunnel vault settlement measured by using RDM method and 3D measurement method are smooth, and the fitting degrees of the curves are greater than 0.95. However, in the aspect of measuring tunnel horizontal convergence, the curve's fitting degree for RDM method is greater than 0.9, and the degree is less than 0.9 for 3D measurement method. So, the measuring accuracy of RDM method is higher than 3D measurement method.More>

2017, 17(3): 56-64.

FullText HTML (142) PDF (547KB) (536)

Effect of surface texture on tribological performance of crankshaft bearing

LIU Cheng, LU: Yan-jun, LI Sha, LIU Wan-wan, YANG Ru

Abstract: The synergistic lubricating effect of grooved texture and dimpled texture was considered, and the compound texture with parabolic grooves and spherical dimples was designed on the surface of crankshaft bearing to improve its lubrication performance. In order to analyze the effect of compound texture on the lubrication performance of crankshaft bearing, a mixed lubrication model of crankshaft bearing was developed based on average Reynolds equation and Greenwood-Tripp micro-convex contact equation, the mass conservation boundary condition was used to deal with the rupture and reformulation of oil film, the tribological performances of crankshaft bearings with grooved texture, dimpled texture and compound groove-dimple texture were analyzed, and the influences of distribution locations and structure parameters of compound groove-dimple texture on the load-carrying capacity and friction force of crankshaft bearing werestudied. Analysis result shows that the compound groove-dimple texture has larger load-carrying capacity than the grooved texture and lower friction force than the dimpled texture. The maximum dimensionless load-carrying capacity is obtained when the optimal groove width is 1.3 mm, the groove area density is 0.7, the maximum groove depth is 25μm, the dimple number is 6, the dimple area density is 0.7, and the maximum dimple depth is 20μm. The minimum dimensionless friction force is also obtained when the optimal groove width is 2.6 mm, the groove area density is 0.7, the maximum groove depth is 30μm, the dimple number is 15, the dimple area density is 0.7, and the maximum dimple depth is 35μm. When the compound groove-dimple texture has optimal distribution location and structural parameters, the load-carrying capacity of textured bearing increases by 4.1% and the friction force reduces by 19.6% compared with the untextured bearing.More>

2017, 17(3): 65-74.

FullText HTML (209) PDF (1399KB) (448)

Mechanics characteristics of surface damage on sliding-rolling friction pair under high-speed and heavy-load condition

PENG Bo, KONG Wen-qin, JIA Lei, LI Yong-jun, WANG Li-qin

Abstract: An elasto-plastic contact model of M50 steel sliding-rolling friction pair was built by using the finite element method. Under the high-speed and heavy-load condition that the contact stress was about 4.0GPa and the velocity was about 50 m·s^-1, the Von Mises stress, shear stress and plastic deformation of surface layer were analyzed, the effects of the friction coefficient and the relative sliding velocity on the contact behaviors of the M50 steel sliding-rolling friction pair were studied, and the plastic deformations of surface layers were compared in M50 steel double-roller sliding-rolling test result. Calculation result shows that the differences of maximum contact stresses and major and minor axis lengths of elliptical contact region between the finite element analysis result and the Hertz theory calculation result are 2.66%, 0.26% and 6.43%, respectively. When the friction coefficient increases from 0.1 to 0.5, the location of maximum Von Mises stress gradually moves from the friction pair subsurface at about 0.5 mm to thecontact surface. When the scuffing failure of friction pair surface occurs, the friction coefficient is more than 0.3, and the maximum Von Mises stress is over 1 700 MPa. When the scuffing failure occurs, the stress and plastic strain of M50 steel friction pair have specific direction. Under the condition that sliding-rolling ratios are 0.12 and 0.15, respectively, the maximum Von Mises stresses are 2 847 MPa and 2 689 MPa on the contact point with higher linear velocity, which leads to relatively large plastic strain, and the maximum values are 0.062 and 0.061, respectively. However, on the surface with lower linear velocity at the two sliding-rolling ratios, the maximum Von Mises stresses are 2 269 MPa and 2 101 MPa, respectively, the plastic strains are smaller, and the maximum values are 0.040 and 0.039, respectively.More>

2017, 17(3): 75-82.

FullText HTML (110) PDF (2801KB) (368)

Influence of elastic element on static and dynamic characteristics of large tilting pad bearing

XI Wen-kui, HAN Qiang-hui, HUANG Tian-hu, XU Jian-ning, JIANG Xiang-jun

Abstract: In order to improve the dynamic stability of warship unit and restrain vibration effectively, a new technology of pivot elasticity was applied to the tilting pad bearing structure of warship propulsion shafting system, and the butterfly spring was mounted at the pad pivot of tilting pad bearing. A large gas turbine was taken as an object, the pivot elasticity structure was introduced into the four tilting pad bearings of large scale shaft, the temperature field, pressure field, stiffness and damping of tilting pad bearing were analyzed by using the fluid-solid-heat coupled model and the multi-field analysis technology, and the effect law of pivot elasticity technology on the tribology and dynamics behavior of tilting pad bearing were investigated. Analysis result shows that when the rotating speed is 3 000 r·min^-1, the maximum oil film pressure of rigid pivot bearing is 6.5 MPa, and the maximum oil film pressure of spring pivotbearing is 6.7 MPa and just increases a little compared with rigid pivot bearing. Meanwhile, the highest temperatures of two kinds of pivot structure bearings are 98.95 ℃ and 98.85 ℃, respectively, so the pivot elasticity technique has little effect on the bearings temperatures. With the increase of rotating speed, the main stiffnesses of two kinds of bearings decrease, but their cross stiffnesses change only in the range of ±0.1MN·m^-1. At the rotating speed of 3 000 r·min^-1, the main stiffness and the main damping of spring pivot bearing are 3.5 GN·m^-1 and 6 MN·s·m^-1, respectively, which are 59% and 39% higher than the values of rigid support bearing. Obviously, the utilization of pivot elasticity technology has little effect on the bearing temperature, the maximum oil film pressure increases slightly, but the main stiffness and main damping of bearing increase obviously, which is favorable for the stability increase and vibration suppression of warship unit.More>

2017, 17(3): 83-89.

FullText HTML (115) PDF (482KB) (359)

Surface texture optimization of journal bearing based on hybrid genetic algorithm

ZHANG Yong-fang, LIU Cheng, LI Sha, LI Xian-wei, YAN Dong, LU Yan-jun

Abstract: The spherical dimple texture was designed on the surface of journal bearing to improve the lubrication performance of crankshaft system in diesel engine. In order to maximize the loadcarrying capacity and minimize the friction factor of the bearing, a hybrid evolutionary optimization method based on the sequential quadratic programming and the genetic algorithm was proposed, and an optimization model was developed for the journal bearing with spherical dimple texture. The distribution location and geometry parameters of dimple texture wereglobally optimized to obtain the optimal angle and maximum depth of texture under given working condition. In order to solve the load-carrying capacity and friction factor of journal bearing, the influence of surface roughness on oil flow was considered, a mass-conservation JFO (Jakobsson, Floberg, Olsson) cavitation algorithm was used to address the rupture and reformulation of oil film, and a mixed lubrication model was developed based on average Reynolds equation and Greenwood-Tripp asperity contact equation. The influence of spherical dimple textures with various distribution locations and geometry parameters (number, area density, and maximum depth) on the load-carrying capacity and friction factor of journal bearing was investigated. Analysis result shows that the load-carrying capacity and friction factor are the monotonic functions of dimple area density. There exists optimal angle and maximum depth of dimple that can maximize the load-carrying capacity and minimize the friction factor. When the eccentricity rises from 0.3 to 0.7, the increment of load-carrying capacity changes from 13.38% to 0.62%, and the decrement of friction factor changes from 0.73% to negative value. Therefore, when the eccentricity is smaller, the spherical dimple texture can increase the load-carrying capacity and decrease the friction factor effectively, and when the eccentricity is larger, the spherical dimple texture is unbeneficial to decrease the friction factor of the bearing.More>

2017, 17(3): 90-98.

FullText HTML (125) PDF (785KB) (422)

Integrated simulation platform of braking system of rolling stock based on multi-discipline collaborative analysis

ZHU Wen-liang, WU Meng-ling, TIAN Chun, ZUO Jian-yong

Abstract: On the basis of the working principle of electro-pneumatic brake, the braking system of rolling stock was taken as research object, one motor car and one trailer were taken as a control unit, and the control subsystem model, air brake subsystem model, electric braking subsystem model and braking executing subsystem model were built based on multi-discipline collaborative analysis. By the correlation parameters of the subsystems, a co-simulation platform was established. The operating condition of full service braking was simulated under the failure of electric braking according to the actual parameters of second phase of Guangfo Metro, and the idling braking time, braking time, braking distance, braking deceleration, instantaneous speed, mean deceleration, longitudinal jerk, coupler force, adhesion utilization and braking cylinderpressure were calculated. The simulation and test results were compared to verify the feasibility and validity of the platform. Simulation and test results show that the braking decelerations of train in simulation and test are about 1.25 m·s^-2 after braking process is steady, the mean braking deceleration in simulation is about 1.05 m·s^-2, the decelerate in test is about 1.09 m·s^-2, all of them meet the requirement that the mean braking deceleration of full service braking is more than 1.0 m·s^-2, and the simulation error is lesser. In full service braking, the adhesion utilizations are different under the equivalent wear braking force distribution strategy, it is 0.13 for motor car and 0.12 for tailor, but both of them are less than the adhesion limited 0.16. Although the qualities of motor car and trailer car are different, the pressures of braking cylinders in full service braking are same under the equivalent wear strategy, and the value is about 420 kPa. In summary, the multi-discipline collaborative analysis and co-simulation platform are effective in modelling and analyzing the braking system of rolling stock, and are theoretical basis to research and develop braking system.More>

2017, 17(3): 99-110.

FullText HTML (162) PDF (1607KB) (538)

Influence of locomotive marshalling mode on air-recharging characteristic of train

HU Yang, WEI Wei, ZHANG Yuan

Abstract: In order to quantify the influence of locomotive marshalling mode on the air-recharging characteristic of heavy-haul train, combining with longitudinal dynamics test results of Shenhua ten-thousand-tons heavy-haul trains, the air-recharging characteristics of train were analyzed, and the train air brake system model was established by using the air brake system simulation method based on airflow theory. The accuracy of the simulation system was verified by test result comparison, and the air-recharging processes on different locomotive marshalling modes, different lag times of multi-locomotives and different pressure reductions were simulated. Calculation result shows that the locomotive's quantity in the front of train has little influence on the air-recharging characteristic of first vehicle, and the difference of auxiliary reservoir pressures for two marshalling modes is less than 15 kPa. The air-charging time of single locomotive marshalling mode is 2.4 times of air-charging time of three locomotives marshalling mode. When the locomotive has centralized on the forepart of the train, the charging time reduction is nonproportional to the increase number of locomotives, which means that the air-charging time of three locomotives concentrated marshalling mode is not three tenth of the air-charging time for single locomotive marshalling train. The influence of locomotive number on air-charging time depends entirely on the marshalling mode. The air-charging time of distributed power marshalling mode with 50 kPa reduction compared to concentrated locomotive marshalling mode decreases by 37%-75%. The air-charging time of locomotive concentrated marshalling mode with 110 kPa reduction is 1.5-3.5 times of distributed power marshalling mode. The air-charging time of distributed power marshalling mode for full service braking is 30%-63% of locomotive concentrated marshalling mode. The lag time of slave control locomotive has little influence on the air-charging time. The increment of air-charging time is similar to the lag time. The quadratic fitting functions of air-charging times of four different locomotive marshalling modes for different decompressions are got, and the air-charging times of the four locomotive marshalling modes increase more slowly with the increase of decompression.More>

2017, 17(3): 111-120.

FullText HTML (129) PDF (757KB) (386)

Review of car-following models of adaptive cruise control

QIN Yan-yan, WANG Hao, WANG Wei, NI Dai-heng

Abstract: The car-following models of adaptive cruise control (ACC) and cooperative adaptive cruise control (CACC) of autonomous vehicles were analyzed. From the aspects of system control principle, vehicle-to-vehicle communication technology and vehicle time-gap, the similarities and differences of ACC and CACC vehicles were expounded. The mainstream car-following models ofACC/CACC vehicles at present were divided into 3 categories: the car-following model based on intelligent drive, the car-following model of PATH laboratory of University of California, Berkeley, and the car-following model based on control theory. The modeling ideas of the 3 categories of car-following models were summarized, and their merits and drawbacks were also expounded. Representative achievements about the impacts of ACC/CACC vehicles on traffic flow characteristics were reviewed from 3 aspects of road capacity, traffic safety, and traffic flow stability. Then, the research status of this field was formed. Meanwhile, the future development trend was pointed out. Research result shows that the impacts of different ACC/CACC carfollowing models on road capacity have relatively big difference. ACC/CACC vehicles are helpful to improve traffic safety. However, because the uniform index of safety evaluation is absent, it is difficult to quantitatively evaluate the impacts of ACC/CACC vehicles on traffic safety. Small scale real vehicle tests validate that ACC vehicles have unstable traffic flow characteristics. The tests reject the numerical simulation-based results of ACC vehicles stability. But both numerical simulations and small scale real vehicle tests show that CACC vehicles can improve traffic flow stability well. Therefore, convincing conclusions are unable to be obtained based on computer simulations absolutely. Real vehicle tests are the necessary way of ACC/CACC research. In order to perfect the research of ACC/CACC in this field, the fundamental diagram models of mixed traffic flow with different proportions of ACC/CACC vehicles, the modeling methods for ACC/CACC car-following models in the intelligent and connected environment and the stability analysis methods for ACC/CACC mixed traffic flow should be built.More>

2017, 17(3): 121-130.

FullText HTML (2183) PDF (592KB) (2023)

Planning model of feeder shipping network for container liners under considering shipper perference

DU Jian, ZHAO Xu, JI Ming-jun

Abstract: To increase the freight demand of feeder shipping network for container liners, a planning model of the network under considering shipper preference was proposed. In the model, the constraints were permitted capacity limit, route operation subsidy and hub-and-spoke cooperation, and the decision variables were route network structure, ship capacity and service frequency. To evaluate the attraction of planning network to shippers with different preferences, the selection proportions of shippers between planning network and existing network were calculated by Logit model after the shipping time and freight of containers were gotten. To solve the model effectively, an intelligent heuristic algorithm was designed, the voyage time, voyage cost, freight per container and freight demand were calculated in the evaluation process ofplanning scheme, and the affiliated ports and sequence of routes were adjusted in the improvement process of planning scheme. The Dalian Port was taken as main port, the 12 ports in Bohai Gulf were taken as feeder ports, and the feeder shipping network for container liners was planned. Calculation result shows that 7 routes are planned among 12 feeder ports. There are 5 208 TEU containers in market, and the freight demand of planned shipping network is 4 420 TEU. The selection proportion of shippers for planning shipping network reaches 85%. When shipper preference is shipping time or cost, the selection proportion of planning shipping network at every feeder port exceeds 60%. Therefore, the planning model of feeder shipping network for container lines under considering shipper preference is effective. The direct routes contribute to attract the shippers with time preference. The multi-anchored routes and higher operation subsidy contribute to attract the shippers with cost preference. Replacing time-window constraint with shipper selection process can improve the optimization effect of the model.More>

2017, 17(3): 131-140.

FullText HTML (113) PDF (714KB) (633)

Sequencing approach of arrival aircrafts based on composite dispatching rules

ZHANG Jun-feng, ZHENG Zhi-xiang, GE Teng-teng

Abstract: To alleviate the flight delay problem, a sequencing approach of arrival aircrafts was proposed based on composite dispatching rules. Based on the machine scheduling theory, the sequencing problem of arrival aircrafts with the target of minimizing weighted total delay was transformed into the machine scheduling problem with the target of minimizing weighted total tardiness. The order-depended time constraint, submitted time constraint and deadline constraint were considered, and the sequencing model of arrival aircrafts was constructed. Through introducing the weighted shortest processing time factor, slack term factor, setup time factor, release time factor and deadline factor, the composite dispatching rule for the sequencing was presented, and a heuristic algorithm for the sequencing was developed. Based on real case, the weighted total delays, total delays and maximum delays of arrival aircrafts computed by using the proposed sequencing approach, first-come-first-service rule and Lingo software were compared. Computation result shows that in the numerical simulation with 30 arrival aircrafts, when the proposed method was compared with the first-come-first-service rule, the weighted total delay reduces by 31 min, and the number of delayed aircrafts decreases by 6. In the northbound operations of Shanghai Pudong Airport, when the proposed method was compared with Lingosoftware, the optimized landing times are same, but 2.4 min per aircraft is saved compared with the actual landing time.More>

2017, 17(3): 141-150.

FullText HTML (130) PDF (924KB) (510)

Traffic sign recognition method based on HOG-Gabor feature fusion and Softmax classifier

LIANG Min-jian, CUI Xiao-yu, SONG Qing-song, ZHAO Xiang-mo

Abstract: In order to improve the accuracy and real-time performance of traffic sign recognition, a traffic sign recognition method was proposed based on HOG-Gabor feature fusion and Softmax classifier. HOG (histogram of oriented gradient) feature was extracted by using the Gamma correction method, and Gabor feature was extracted by using the contrast limited adaptive histogram equalization method. According to the linear feature fusion principle, HOG and Gabor feature vectors were connected to constitute the fusional feature vector for depicting the traffic signs. Theeffectiveness of the proposed method was verified based on the GTSRB (German Traffic Sign Recognition Benchmark) data set. The recognition effects of traffic sign based on single feature and fusional feature were compared. Experimental result shows that in image enhancement, the classification accuracy based on HOG feature is 97.11% and is largest by the Gamma correction method, and the classification accuracy based on Gabor feature is 97.54% and is largest by the contrast limited adaptive histogram equalization method. The minimum classification accuracy is 97.11% by using Softmax classifier, and classification time is only 2 s. The correct recognition rate of traffic sign reaches 97.68% by using the proposed method based on HOG-Gabor fusional features, so the traffic sign recognition method based on HOG-Gabor fusional features and Softmax classifier has high recognition rate and real-time performance.More>

2017, 17(3): 151-158.

FullText HTML (405) PDF (2007KB) (627)

User Center

Journal Info

Collection of Source JournalsMore>

2017 Vol. 17, No. 3

Search Type

Search Journal

LinksMore>