交通运输工程学报

Hot spot SCF computation method of concrete-filled and PBL-stiffened rectangular hollow section joint subjected to axial tensions

LIU Yong-jian, JIANG Lei, XIONG Zhi-hua, ZHANG Guo-jing, Amir FAM

Abstract: The width ratio and thickness ratio of brace to chord and the width-to-thickness ratio of chord for concrete-filled and PBL-stiffened rectangular hollow section joint subjected to axial tensions were considered, and the finite element model of hot spot stress concentration factor (SCF) was established.The computation result of hot spot SCF was fitted by the least square method, the SCF computation formulas under different geometric parameters were proposed, andthe hot spot SCFs and load ranges of rectangular hollow section joint and concrete-filled rectangular hollow section joint stiffened with PBLs were compared by using the proposed formulas.Computation result shows that the SCF curve calculated by using the finite element model is almost consistent with the experiment curve obtained by the static test, and the average ratios of finite element calculation results to CIDECT calculation results are 1.006, 1.007, 1.013, 1.015 and 0.987 at the hot spots at the joint of brace and chord, respectively, and the differences are less than 15%, which verifies the reliability of finite element model.The SCFs of concrete-filled and PBL-stiffened rectangular hollow section joint subjected to axial tensions have the similar variation trend and change in parabola shape with the width ratio of brace to chord.The maximum value of SCF appears when the width of brace to chord is between 0.6 and 0.8, and increases when the width-to-thickness ratio of chord and the thickness ratio of brace to chord increase, which is same with the SCF of rectangular hollow section joint calculated by CIDECT.The computation formulas and finite element model of hot spot SCF of concrete-filled and PBLstiffened rectangular hollow section joint subjected to axial tensions are compared, the SCF average ratio is 1.011, the mean variance is 0.222, and the variation coefficient is 0.219, which proves that the fitting formulas are accurate and reliable.Concrete-filled and PBL-stiffened rectangular hollow section joint is compared with rectangular hollow section joint, the SCFs of brace and chord computed by using the proposed formulas decrease by more than 68% and 61%, respectively, and the allowable load ranges increase to more than three times under the action of 2.0×10⁶ cycle times.More>

2017, 17(5): 1-15.

FullText HTML (150) PDF (1927KB) (1040)

Calculation method of shear bearing capacity of R-UHPC beam

MA Xi-lun, CHEN Bao-chun, YANG Yan, HUANG Qing-wei, SU Jia-zhan, WU Qian-wen, SHEN Xiu-jiang

Abstract: A lot of test data of shear bearing capacity concerning reinforced ultra-high performance concrete (R-UHPC) beams were collected, existing calculation methods of shear bearing capacity were analyzed, and the shear bearing mechanism of R-UHPC beam was studied.The tensile strength of UHPC was considered, a calculation method of shear bearing capacity for R-UHPC beam based on truss-arch model was presented, and the calculated result and test result of shearbearing capacity were compared.Comparison result indicates that in the calculation methods at present, based on statistical analysis method, the average ratio of the calculation values to the test values of shear bearing capacity is 0.92, its standard deviation, correlation coefficient and reliability coefficient are 0.23, 0.78 and 0.877, respectively, and the computation accuracy is not high due to the limitation of regression data.For the calculation methods based on the general truss model, the average ratio of the calculation values to the test values of shear bearing capacity with AFGC in French UHPC is 0.90, its standard deviation, correlation coefficient and reliability coefficient are 0.18, 0.80 and 0.891, respectively, and the computation accuracy is higher than the accuracies of Japan UHPC's JSCE and Swiss SIA. When the effect of longitudinal reinforcement is considered based on AFGC, the average ratio of the calculation values to the test values of shear bearing capacity is 0.93, its standard deviation, correlation coefficient and reliability coefficient are 0.23, 0.75 and 0.858, respectively, and the result has larger discreteness than the result of AFGC.For the calculation methods based on the truss-arch model, the average ratio of the calculation values to the test values of shear bearing capacity is 0.76, and its standard deviation, correlation coefficient and reliability coefficient are 0.26, 0.62 and 0.768, respectively.Because the computation methods of shear bearing capacity of reinforced (common) concrete beam are straightly applied and the tensile strength of UHPC is neglected, the result is conservative and most unreliable.For the proposed calculation method of shear bearing capacity, the average ratio of the calculation values to the test values of shear bearing capacity is 0.94, and its standard deviation, correlation coefficient and reliability coefficient are 0.21, 0.80 and 0.885, respectively.Compared with the calculation methods at present, the proposed method has higher precision and less discreteness.More>

2017, 17(5): 16-26.

FullText HTML (264) PDF (665KB) (799)

Calculating method of structural robustness of double-tower cable-stayed bridge with steel truss girder

ZHENG Xiao-bo, ZHAO Yu, HE Shuan-hai, ZHANG Gang

Abstract: In order to ensure the structural robustness of double-tower cable-stayed bridge with steel truss girder, all-round multipoint temperature compensation system was developed based on the Xinjiang Guozigou Bridge and field structure test.The strain, deflection and cable force increment of steel truss girder were measured under specific loading conditions, and the specific locations of important components of steel truss girder and stay cables were determined.Based on the experimental result and the generalized structural stiffness theory, the robustness importance factors of cables and chords were analyzed by using the finite element model of bridge structure, and the most critical failure model of cable-stayed bridge was studied.Study result shows that the regularities of the measured data of steel truss girder strain under various working conditions are good, and the ratios of measured and theoretical values of strain and deflection are less than 1.0, so the bearing capacity and anti-deformability of steel truss girder meet the design requirements, and the girder has sufficient safety reserve.The maximum deflection of the girder under eachloading case is at the middle of main span and can reach to 237 mm, so the girder has strong antideformability.Because the ratios of the measured increments of cable force and theoretical values are less than 1.0, the stay cables have safety reserve to a certain extent.The chords at control sections and the specific stay cables of the girder have higher importance factors, the factors of stay cables are larger than the values of the chords, and the importance factor distributions of the chords concentrate in the middle of main span and nearby areas of main pylons.According to the distributions of importance factors of stay cables, the damage of single stay cable can causes the collapse of overall structure, and the failures of more than two stay cables will result in the progressive collapse of entire structure.The failures of the longest stay cable of main span and the chords at its middle, as well as the chords at the supports of side spans and near the main towers, are more adverse to the entire bridge.More>

2017, 17(5): 27-38.

FullText HTML (150) PDF (1431KB) (1)

Computational method of 3D aggregate angularity based on CT images

CUI Zhe, ZHANG Sheng-rui

Abstract: To improve the evaluation accuracy of aggregate angularity, a computational method of3 D aggregate angularity was put forward.Based on CT technology and 3 D reconstruction technology, image intensification and sharpen filter were used to clearly display the aggregates in aggregate CT images, and the gray thresholds of enhanced CT images were calculated and divided.MIMICS was used to reconstruct the 3 D models of aggregates in Marshal specimens.The roughness degree and sphericity degree of aggregate were put forward and taken as evaluation indexes to evaluate aggregate angularity based on the 3 D model.The influence factorsof 3 D model reconstruction were analyzed.Computational result shows the gray values of aggregate, asphalt and pore in AC-16 Marshall specimens are 101.32-170.14, 4.32-101.32, and0-4.32, respectively, so, image intensification and sharpen filter can clearly display aggregates in CT images and improve the 3 Dreconstruction accuracy of aggregate.The computational standard deviation of sphericity degrees is 0.000 7 when the CT images are dealt with by 2 pixels×2 pixels and 3 pixels×3 pixels sharpen filters, and the standard deviation of sphericity degrees is 0.042 3 when the CT images are dealt with by 5 pixels×5 pixels, 6 pixels×6 pixels and 7 pixels×7 pixels sharpen filters.Therefore, 2 pixels×2 pixels or 3 pixels×3 pixels sharpen filter should be adopted to dealt with the CT images in order to ensure the low fluctuation of computed sphericity degree.When sampling point densities are 50 and 70 per cubic millimeter, the standard deviation of roughness degrees is 0.001 6, but when sampling point densities are 5, 15 and 25 per cubic millimeter, the standard deviation of roughness degrees is 0.034 9, so, the density range of sampling points in CT images should be 50-70 per cubic millimeter to ensure that the 3 D models can reflect the real shapes of aggregates precisely.The standard deviations of 2 D sphericity degrees and roughness degrees based on 5 CT cross sections are 0.012 1-0.048 2, which reflects the larger variability and deviation.However, the roughness degrees of aggregates 1-5 calculated by the method based on the 3 D models are 0.991 2, 1.032 1, 0.974 2, 1.075 1, 1.043 2, respectively, and the corresponding average 2 Droughness degrees of aggregates 1-5 are 0.994 1, 1.023 9, 0.988 3, 1.097 5 and 1.060 8, respectively, which shows the 3 Dand 2 Droughness degrees are basically same.Obviously, the computational method based on the 3 D model used to calculate the roughness degree and sphericity degree fully considers the 3 D angularity of aggregate, the calculation result is not affected by the selected CT cross sections, and there are no calculation variability and deviation.More>

2017, 17(5): 39-49.

FullText HTML (195) PDF (4137KB) (484)

Properties of fireproofing coating with dirt resistance and water resistance on tunnel surface

JING Hong-jun, CUI Shi-fu, AI Tao, JING Hong-wei

Abstract: In order to improve the dirt resistance and water resistance of fireproofing coating of tunnel surface, waterborne epoxy resin was taken as adhesive, organic silicon polymer, nanometer silicon dioxide and titanium dioxide were taken as antifoulants, brucite fibre was taken as inorganic flame retardant, ammonium polyphosphate (APP) /melamine (MEL) /pentaerythrotol (PER) was taken as organic flame retardant system, the appropriate proportions of organic flameretardant system's components were determined by thermo analysis, the effects of nanometer silicon dioxide, nanometer titanium dioxide, and organic silicon against on the dirt resistance of fireproofing coating were studied by fire test, thermo-gravimetric analysis, sticky dirty resistance test and water resistance test, and the excellent new-type multifunctional fireproofing coating with dirt resistance and water resistance was developed by screening and optimization experiments of formula, fire endurance test and microscopic analysis.Analysis result shows that when the mass ratio of APP, PER and MEL is 2∶1∶1, the charring temperature peak of epoxy resin decreases from 331 ℃to 277 ℃ under the catalysis of APP on PER, and the carbon residue rate of epoxy resin increases by 10%in the temperature range of 270 ℃-800 ℃.The charring peak temperature of APP-PER is lower about 66 ℃ than the decomposition temperature of MEL and about 100 ℃ than the charring temperature of epoxy resin degree, which is helpful to form expanding carbonaceous layer with good heat insulation effect.When 5% organic silane or 3%nanometer silica and 5% nanometer titanium dioxide is added into the flame retardant system comprised of APP, PER and MEL, the fireproofing time of coating is 60 min or 80 min.When quantitative organic silane, nanometer silicon dioxide and nanometer titanium dioxide are added into the flame retardant system together, the fireproofing time of coating increases to 100 min.Then 10%zinc stannate is added into the system, the fireproofing time of coating increases to 120 min.The increase of fireproofing time shows the significant coordination fireproofing effect of the components.The final formula of optimized intumescent fireproofing coating is that the mass ratio of waterborne epoxy resin, silane, composite flame retardant, curing agent, composite nanometer powder and zinc stannate is 40∶5∶33∶4∶8∶10, its dirt resistance grade is 1, its fireproofing time is more than 120 min, and there are no peeling and cracking phenomena after800 himmersion.More>

2017, 17(5): 50-60.

FullText HTML (173) PDF (1003KB) (556)

Characteristic of wind-sand flow field of box-type movable sand barrier

CHENG Jian-jun, ZHI Ling-yan, XUE Chun-xiao

Abstract: On the basis of field investigation, the sand-controlled characteristics of box-type movable sand barrier and the evolution process of wind-sand flow field around the barrier were analyzed by using computational fluid dynamics three-dimensional numerical modeling method and indoor wind tunnel test under the effect of changeable porosity and wind velocity.Analysis result indicates that under the control of sand barrier, there are deceleration zone, deceleration uptrend zone, acceleration zone, and vortex zone successively before and after sand barrier along the direction of wind.Deceleration zone and vortex zone form obviously in the cavities of sand barrier, the transient wind-sand flow sinks down in deceleration zone, vortex zone behind and the cavities of sand barrier, so the box-type movable sand barrier plays an important role in sandcontrol.As the increase of wind velocity, the height of deceleration uptrend zone increases, but the porosity change of windward side of sand barrier does not cause any effect to the height of deceleration uptrend zone.When the porosity of windward side of sand barrier is less than 5%, the flow weakens observably, and the sand deposites in the front of sand barrier, which can not make full use of the sand-control function of vortex zone behind sand barrier.When the porosity is more than 25%, the sand barrier can play full sand-controlled role, sand deposites in thewindward side, leeward side and cavities of sand barrier.When the porosity continues to increase to 30%, the sand-controlled effect does not improve obviously.Under the condition that the porosity is 30%, when the wind velocity continues to increase, the sand accumulation increases behind the sand barrier and decreases in the cavities of sand barrier, however, the tendency of decrease after the first increase occurs in the windward side of sand barrier.More>

2017, 17(5): 61-70.

FullText HTML (171) PDF (3533KB) (303)

Simplified calculation method of critical value of 3D derailment coefficient under quasi-static condition

WANG Jian, MA Xiao-chuan, MA Dao-lin, WANG Ping

Abstract: According to the coordinate system transformation relationship in wheel-rail system, the 3 Dstress analysis model of wheel-rail contact spot was established under the quasi-static condition, the formula considering wheelset yaw angle and wheel-rail creep force for calculating3 D derailment coefficient was derived, and the calculation method of critical value of 3 D derailment coefficient was obtained when the wheel was in the critical state of derailment.Taking the LMA wheel tread and CHN60 rail profile as examples, the influence rule of wheelset yawangle and friction coefficient on the critical value of 3 Dderailment coefficient was analyzed and compared with the critical value of Nadal derailment coefficient.To simplify the calculation method of 3 Dderailment coefficient, the ratio relation between the Kalker linear synthetic creep force and the three times of Coulomb friction force was discussed according to Shen-HedrickElkins creep model under different wheelset yaw angles, friction coefficients and vertical forces.Through the analysis of variation rules of the ratios between lateral and longitudinal creep forces with different wheelset yaw angles and friction coefficients, a simplified calculation method of 3 D derailment coefficient was proposed under the quasi-static condition and compared with the exact formula.Analysis result shows that compared with the threshold of 3 Dderailment coefficient, when the wheelset yaw angle is less than 1.5°, the proportion of longitudinal creep force in the tangential force is significantly greater than that of lateral creep force, which causes the threshold of Nadal derailment coefficient to be more conservative.However, when the wheelset yaw angle is larger, the proportion of lateral creep force in the tangential force is more than 90%, and the calculated critical values of 3 Dand Nadal derailment coefficient are basically same.In addition, the wheel-rail contact spot has reached purely sliding state in the critical state of wheel derailment.The ratio between lateral and longitudinal creep force is not affected by the friction coefficient and has a strong linear relation with the wheelset yaw angle.Compared with the exact formula, the error by using the simplified method is within±5%, and the simplified method can meet the requirement of engineering application.More>

2017, 17(5): 71-80.

FullText HTML (151) PDF (927KB) (759)

Research on wheel wear under locomotive traction condition

YANG Yang, DING Jun-jun, LI Fei, LI Dong-yu, LI Jin-cheng

Abstract: Taking a running C0-C0 shaft type electric locomotive as research object, the wheel wear calculation model of electric locomotive was established based on the Archard wear model, and the influence of locomotive drive system was considered in the model.The wheel wears under constant speed and starting conditions were studied.The wheel wear was calculated according to a actual line, and was compared with the measured data.The abnormal wear of wheel flange during the normal operation of locomotive was studied.Analysis result shows that when the vehicle runs 2.6×10⁵ km at a constant speed, and the traction forces increase from 40 kN to120 kN and 120 kN to 200 kN, the wears increase by 0.74 mm and 1.74 mm respectively, so the wear increases rapidly with the increase of traction force.Increasing the traction force during locomotive starting process can obtain greater acceleration.With the increase of traction force, the creep rate increases obviously.Increasing the traction force can save running time, but wear increases at the same time.Compared with the measured data of wheel wear, the wheel wear calculation model is comparatively accurate, and the simulation result at the tread surface is ingood consistency with the measured result.Because the plastic flow of materials and the influence of railway switch are not considered in the wheel wear calculation model, the simulation result of wheel rim has difference with the measured result.The wheel wear greatly reduces when reducing the transverse momentum of second wheelset and rail-side lubrication. When the transverse momentum of second wheelset reduces from 15 mm to 10 mm, the cumulative wear of middle wheelset reduces by 15.4%.After rail-side lubrication, the maximum cumulative wears of the first wheelset, the second wheelset and the third wheelset decrease by 13.40%, 21.32%, 6.46%, separately.More>

2017, 17(5): 81-89.

FullText HTML (199) PDF (1593KB) (372)

Numerical simulation of critical hydroplaning speed of aircraft tire under wet pavement condition

LI Yue, CAI Jing, ZONG Yi-ming

Abstract: A fluid-solid coupling analysis model of aircraft tire and wet pavement based on CEL algorithm was developed by using ABAQUS.The expressions of hydrodynamic pressure of tire contact and vertical supporting force of pavement were derived.The taxiing conditions between aircraft take-off and landing process were compared.The concepts of upper and lower limit solutions of critical hydroplaning speed were proposed.The features of static deformation and dynamic hydroplaning of tire model were verified.The influence rules of tire pressure, tire pattern and water-film thickness were discussed. The contact area and distribution of hydrodynamic pressure for tire were analyzed.Simulation result indicates that the tire contact area is0.076 m2 under axle load of 76.6 kN, the vertical deformation at the centre of tire is 3.27 cm, and thecritical hydroplaning speed is 128.5-222.4 km·h^-1, which is in consistence with the result of NASA's tire hydroplaning test.Therefore, the rationality and feasibility of simulation model are proved.When tire pressure is 1 140 kPa, the critical hydroplaning speed of aircraft tire under decelerating impact is 163 km·h^-1 and lower than upper limit of accelerating impact (226 km·h^-1), the tire contact area obviously reduces, and the supporting force from the pavement to the tire is less than 10% of wheel load.In comparison with the calculation result of NASA's empirical equation, the lower limits of critical hydroplaning speed under decelerating impact are more conservative within the scope of tire pressure from 450 kPa to 1 109 kPa, and the difference is30-70 km·h^-1.The drainage effect of radial tire pattern can reduce the peak value of hydrodynamic pressure at the leading edge of aircraft tire and increase the tire contact area.The critical hydroplaning speed of aircraft tire with tire pattern under decelerating impact increases by26.9%-28.8% comparing with the speed of smooth tire, and the amplification is twice as much as that of accelerating impact.As the water-film thickness increases from 3 mm to 13 mm, the upper and lower limits of critical hydroplaning speed of aircraft tire respectively reduce by 85 km·h^-1 and 43 km·h^-1 when the tire pressure is 1 140 kPa.In case of lower tire pressure and thicker water-film, the lower limit of critical hydroplaning speed is merely 127 km·h^-1 under decelerating impact and lower than most aircrafts'landing speeds 205-250 km·h^-1, so the risk of hydroplaning accident increases.More>

2017, 17(5): 90-101.

FullText HTML (206) PDF (2653KB) (601)

Numerical simulation of righting process for damaged-capsized hull

PAN De-wei, LIN Cheng-xin, ZHOU Zhao-xin, SUN Yu-qiang, LIU Zhi-jie

Abstract: The buoyancy and stability of damaged-capsized hull were considered, and the spatial position and mechanical of hull state were studied during righting process.The spatial mechanical equilibrium equation of hull was established by Euler rotation transformation method.The stability and righting mathematical model of hull were derived by using the hydrostatical theory of ship.The flooding quantity was calculated according to Bernoulli theorem and its impact on the positions of barycenter and buoyant centre were obtained.The righting process of damagedcapsized hull was simulated by using General HydroStatics (GHS) software, the maximum righting force and flooding quantity were solved, and the shear force, bending moment, torque of six longitudinal positions along the hull were calculated.Computation result shows that the righting moment of damaged-capsized hull decreases by 130.312 MN·m in the early righting process because water floods damaged cabins.So, the flooding water decreases the stability ofdamaged-capsized hull and the righting force, which results in the decrease of righting difficulty.In the later righting process, the maximum tilting moment of damaged-capsized hull is163.594 MN·m because water floods damaged cabins.Thus, flooding water decreases the stability of hull, increases the righting difficulty, and the larger righting force is needed to balance the hull.The longitudinal strength distribution of hull changes in response to the righting force and flooding quantity.The maximum force and torque of multi-point righting are less than40% and 50% of the corresponding values of single-point righting, respectively.The maximum flooding quantities in Schemes 1-4 are 6 269.76, 6 781.01, 5 830.76 and 6 653.33 t, respectively, which shows that the flooding quantities of single-point righting (Schemes 1-3) are less than the flooding quantity of multi-point righting (Scheme 4) through reaonably arranging the positions of righting points.More>

2017, 17(5): 102-112.

FullText HTML (171) PDF (2803KB) (392)

Passenger evacuation model of metro platform considering psychological stress

CHEN Shao-kuan, DI Yue, LI Fang, XU Qi, LIU Hai-dong

Abstract: The effects of psychological stress from emergencies on metro platform on the evacuation behaviors of passengers were analyzed. The numerical relationship between psychological stress and retention time, local density, distance of hazard was built.The expected velocity of evacuation under psychological stress was researched, a modified expectation velocity model was developed, and the social force model was modified.Aiming at the evacuation process of passengers on the platform, a multi-agent perception and decision-making flow was designed.An evacuation simulation model based on multi-agent technology was built, and the whole process from the reception of evacuation signals to the evacuation completion for passengers was abstractly dealt with.AnyLogic simulation software was used to build the simulation case, the built models were imported into the simulation software by interface means, and the evacuationbehaviors of passengers were studied based on the Xizhimen Platform in Beijing Metro Line 2.Simulation result shows that the psychological stress has large effect on the evacuation time, evacuation velocity, passenger density and detour distance.The passengers'evacuation efficiency with psychological stress increases by 24.03%than that without psychological stress, the average total evacuation time reduces by 27.68%, and the average evacuation times in stair region and connecting region reduce by 36.20% and 22.05%, respectively.The passengers'average evacuation velocity with psychological stress in stair region is about 0.226 m·s^-1, and the velocity without psychological stress is 0.351 m ·s^-1.The highest gathering density in connecting region with psychological stress is 8.0 passenger· m^-2, while the density is9.5 passenger·m^-2 without psychological stress.The highest gathering density in stair region is3.5 passenger·m^-2.The detour distance with psychological stress increases by 96.91% than that without psychological stress, and the movement of passengers with psychological stress is more chaotic.More>

2017, 17(5): 113-120.

FullText HTML (221) PDF (1230KB) (785)

Review of pavement detection technology

MA Jian, ZHAO Xiang-mo, HE Shuan-hai, SONG Hong-xun, ZHAO Yu, SONG Huan-sheng, CHENG Lei, WANG Jian-feng, YUAN Zhuo-ya, HUANG Fu-wei, ZHANG Jian, YANG Lan

Abstract: The important research results of pavement detection were summarized. The development state of technology detecting the damage, roughness, rutting, skid resistance (structure depth) and structural strength (deflection) of pavement was analyzed.The shortage and development direction of pavement detection technology were studied.Research result indicates that the development of pavement detection technology at home and abroad hasexperienced three stages: early traditional manual detection, semi-automatic detection at the end of 20 th century and nondestructive automatic detection at present.The main features of nondestructive automatic detection are rapidness and intellectualization, and the damage, roughness, rutting, skid resistance and structural strength of pavement as well as the alignments and facilities of roads are detected simultaneously because multi-source sensors work together and are integrated in the multi-purpose road test car.In term of pavement damage detection, digital image detection technology is used for the rapid detection of pavement cracks.In term of pavement roughness detection, laser displacement sensing technology is used to realize fast automatic detection.In term of pavement rutting detection, laser and digital image technology are used to realize noncontact intelligent detection.In terms of skid resistance detection and structural strength detection of pavement, the correlativity between the results detected by using sand paving method and Beckman beam method is established to realize the rapid detection of structure depth and deflection of pavement based on laser technology.In order to reduce the interference of external factors to the existing detection technology and detection equipment, and to improve the signal to noise ratio of detection, the road detection and data processing methods suitable for various working conditions should be developed to realize the high efficiency and intellectualization of pavement detection.More>

2017, 17(5): 121-137.

FullText HTML (3522) PDF (891KB) (2236)

Design method of active suspension LQG controller for rolling motion safety of vehicle

WANG Ya-xiong, CAI Yu-meng, WANG Jian, YAO Ming, CHEN Shi-an

Abstract: To improve the steering-rolling motion safety of vehicle, a LQG controller for active suspension was developed.A dynamics model with 3 degrees of freedom depicting vehicle steering-rolling motion was established.Lateral-load transfer ratio (LTR), rolling angle and its acceleration were chosen to construct the evaluating indexes of comprehensive performance for vehicle rolling motion safety.In order to solve the problem that the general design approachs couldn't track the system disturbance term, namely the steered angle of front axle, and the control vector couldn't be solved because the determinant of weighted control coefficient matrix equaled to zero, the new method was scheduled as follows: the steered angle of front axle was introduced into a differential equation in accord with the minimum phase system, the differential equation was combined with the original system equation into an augmented system equation, andthe infinitely small terms including control terms were added to the evaluating indexes of comprehensive performance to meet the design condition of LQG controller.Based on the analytic hierarchy process (AHP) and normalization method, the statistical data of automotive steering-rolling motion were obtained by simulation under fish-hook working condition, and then the weighted coefficients of LQG controller were determined.The working effects of LQG controller of active suspension for steering-rolling motion safety of vehicle were verified by numerical simulation under multiple working conditions.Simulation result shows that the developed LQG controller does not interfere with the steering operation of driver.Compared with the passive suspension, when the active suspension is regulated by the developed LQG controller under fish-hook working condition, slalom working condition and double-lane change working condition, the variances of LTR, namely the most important evaluating index of rolling motion safety, decrease by 32.08%, 32.82%, and 29.24%, respectively, the variances of rolling angle reduce by 47.74%, 44.19%, and 63.41%, respectively, and the variances of rolling angle acceleration reduce by 87.30%, 60.00%, and 86.39%, respectively.So, adopting the active suspension with the new developed LQG controller can significantly improve the rolling motion safety of vehicle and achieve good steering condition adaptability.More>

2017, 17(5): 138-148.

FullText HTML (115) PDF (696KB) (429)

Conformal detection algorithm of anomalous behaviors of vessel

MA Wen-yao, WU Zhao-lin, LI Wei-feng

Abstract: In order to accurately detect the anomalous maneuvering behaviors of ship and reduce the false alarm rate of anomalous behaviors, a conformal detection algorithm of anomalous behaviors was proposed.Some characteristics was introduced into the trajectory points of ship to reflect the maneuvering modes, and the similarity of maneuvering behavior was measured through altering course behavior and speed-changing behavior.The integrated behavior similarity of ship was defined by combining the spatial location similarity and the maneuvering behavior similarity, the nearest neighbor feature points of single track points on training trajectory sequence were calculated, and the conformal detection sample sequence was constructed.In order to overcome the sample overlapping situation of class distribution, the singular value measure of conformal detection algorithm was improved, the nonconformance score between the samples was calculated by the integrated behavior similarity, and the randomness test value of single track point wasused to determine the distribution conformance of trajectory point and sample sequence.The real AIS data of Qiongzhou Strait were taken as the normal data, the random anomalous trajectories were simulated by the computer, the artificial abnormal maneuvering behaviors were defined, and the abnormal detection tests were carried out.Experimental result shows that the detecting accuracy rate of random anomalous behaviors is 100%, but a part of normal trajectories in the set of evaluation trajectories are divided into anomalous trajectories by mistake.When the confidence levels are 99.0% and 99.7%, respectively, the false alarm rates are 0.6% and 0.2%, respectively, and less than the significance levels of 0.01 and 0.003, respectively.Therefore, the algorithm can effectively detect the random abnormal trajectory generated by the computer, strictly control the false alarm rate by using designated significance level, and detect artificial changing speed and altering course abnormal behaviors, and the test result changes with the change of ship behavior.More>

2017, 17(5): 149-158.

FullText HTML (206) PDF (1655KB) (1174)

User Center

Journal Info

Collection of Source JournalsMore>

2017 Vol. 17, No. 5

Search Type

Search Journal

LinksMore>