Vehicle trajectory data contains massive spatial-temporal traffic information, which is one of the necessary data for traffic state estimation. To solve the problem that it's difficult to obtain the fully sampled vehicular trajectory in the existing data collection environment, oriented to the connected and automated environment, a fully sampled trajectory reconstruction model of mixed traffic flow is proposed . Firstly, vehicle composition and trajectory data collection environment of mixed traffic flow with the connected automated vehicle (CAV) are analyzed. Then, a vehicle trajectory reconstruction model is proposed based on intelligent driver car-following model. Based on this, the number of inserted trajectories, trajectory position and speed are estimated. Finally, numerical simulation is designed to investigate the influence of traffic density and penetration rate of CAVs. Results show that, when the penetration rates of CAV and connected vehicle (CV) at 8% and 20%, respectively, the model can reconstruct more than 84% vehicular trajectories under different traffic densities. The accuracy of the reconstructed trajectories increases with the increase in penetration rates of CAV and CV. Besides, when the traffic density is 70 veh/km and the penetration rate of CAV at a low level of 4%, the proposed model can reconstruct 82% vehicular trajectories.

To clarify the vehicle trajectory characteristics and driving behavior preference of hairpin curves of mountain roads, the vehicle trajectory and lateral distance between wheel track and lane marking on hairpin curves were collected in the condition of natural driving, the track curvature were calculated with the measured track data, the relationship between the track curvature and road design curvature were analyzed, and the changing mode of the track curvature was determined. Moreover, the concept of track equivalent radius was presented, and the cornering behavior and typical track patterns on hairpin curves were analyzed. It is found that 1) serious lateral deviation can be observed at entrance, middle and exit of a hairpin curve. 2) At the entrance of a hairpin curve, the vehicle has begun the state of curve driving before running on the transition section, and at the exit, the track curvature reduces to zero on a tangent after departing the transition section; the change rate of track curvature is lower than that of the transition. Gradient of the left-turn track is lower than that of the right-turn track. 3) The curvature of the left-turn track at the middle of the hairpin curve is lower than or close to the design curvature, while the curvature of the right-turn track is higher than the design curvature. 4) The equivalent radius of left-turn track is higher than the design radius of the hairpin curve, while the minimum and average equivalent radius of the right-turn track are generally lower than the design radius. 5) Drivers can increase and maximize the track radius with different cornering behaviors, but need to drive on the opposite lane. 6) When a driver cuts a corner, the increment in track radius of the left-turn curve is higher than that of the right-turn curve; namely, it is easier for a vehicle to turn left into the hairpin curve to cut a corner; of three types of hairpin curves, big, flat and small curves, which correspond to the deflection angles larger than 180°, equal to 180°, and less than 180°, the curves with the deflection angles larger and less than 180° lead to better vehicle cornering performance.

Combining autonomous vehicle (AV) with time-sharing leasing or peer-to-peer (P2P) leasing, two novel travel modes are provided for travelers. To explore the behavior characteristics of AV rental, travelers’ preferences and influence factors for choosing AV time-sharing leasing, AV P2P leasing, private cars and public transit are analyzed. Based on the preference survey data of travel mode choice, the structural equation-multinomial Logit (SE-MNL) model combining structural equation model (SEM) with multinomial Logit (MNL) model is established, which calibrates the parameters of the observed and latent variables. The comparative analyses are done for the parameter calibration results of the MNL and SE-MNL models. The results show that at 95% confidence level, both the observed variables of travel cost, in-vehicle time, driver’s license status, travel purpose, marital status and the latent variables of convenience, safety, ride experience, ride comfort have significant effects on travelers’ choice of AV time-sharing leasing or P2P leasing. The goodness of fit of the SE-MNL model is 2%～3% higher than that of the MNL model.

Autonomous vehicles have an important influence on travel mode choices of travelers, which can affect urban traffic demand, urban spatial layout and urban planning. Based on the extended technology acceptance model and the heterogeneity of different travelers’ preferences, a random parameter Logit model (RPLM) with latent variables is established to study the impact of travel characteristics, psychological latent variables and individual socio-economic attributes on the travelers’ behavior of choosing autonomous vehicles. The results show that, RPLM has a better fitting degree than the traditional multiple Logit model. Different travelers have heterogeneous preferences for travel cost, and the travel cost follows a normal distribution in the utility function. Travelers’ choice behavior of autonomous vehicles is not only affected by travel characteristics and socioeconomic attributes, but also by psychological latent variables, such as perceived trust, social norms and behavioral intention. Reducing the travel cost of autonomous vehicle can significantly increase the probability that travelers choose autonomous vehicle.

In order to study the seismic performance of assembled concrete piers in different connection modes, the quasi-static test of two assembled concrete piers (using steel tube shear connector and grouting sleeve connection, respectively) and a cast-in-place integral concrete pier was conducted, the hysteresis curve, skeleton curve, ductility, stiffness degradation and energy consumption capacity of the comparison test specimens were analyzed, the finite element model was established by ABAQUS general procedure ,and the finite element parameters were analyzed. The results show that the hysteretic loops of the three types of piers are full and have good seismic performance, the failure mode of specimens is integral bending failure, there is no obvious strength degradation, and the piers have similar energy dissipation capabilities. Under the conditions of different axial compression ratio, slenderness ratio, concrete strength and steel strength, the horizontal peak load and displacement ductility coefficients of assembled concrete piers with steel tube shear connector are better than that connected with traditional grouting sleeves, and the increase amplitude is 4%−32% and 8%−36%, respectively. Axial compression ratio, slenderness ratio, embedding depth of steel tube shear connector and steel tube diameter are important parameters that affect the seismic performance of assembled concrete piers connected by steel tube shear connector.

In order to find a nondestructive and quantitative monitoring method for steel corrosion in reinforced concrete, theoretical formulas of the partial modulus and corrosion rate of rebar are derived with consideration of the influence of the Earth’s background magnetic field and environmental interference magnetic field. Accelerated corrosion tests were performed on 8 steel bars through electrification to obtain specimens with different degrees of corrosion. A rebar corrosion monitoring system was then developed to measure the magnetic field intensity of rebar before and after corrosion. Finally, the partial modulus of the magnetic gradient tensor is used to calculate the corrosion rate of the rebar. Results show that the magnetic field intensity curve of the rebar is generally shifted after corrosion, but the absolute value after corrosion may increase or decrease compared with that before corrosion, without consistency. The absolute value of the magnetic gradient and the average value of the partial modulus of the steel bar after corrosion are less than their counterparts before corrosion. In the magnetic monitoring of rebar corrosion, the magnetic gradient and partial modulus of the rebar are much larger than those of the environmental magnetic field, and therefore the gradient of the environmental magnetic field and its partial modulus are negligible. The minimum error between calculated and measured corrosion rates of the 8 specimens is 0.22%, while the maximum error is 9.40%, and the average error is 3.92%, with a standard deviation of 3.32%.

In order to investigate the effect of geogrid on the mechanical properties of clay-contaminated ballast, a series of direct shear tests were carried out on geogrid-reinforced ballast specimens under four normal pressures and three degrees of clay pollution. The effects of geogrid on the shear strength and shear deformation of ballast specimens were compared and analyzed. The reinforcement effect of geogrid on ballast with consideration of clay pollution was studied. The results show that the geogrid can increase the shear strength of clean and clay-contaminated ballast samples, and the peak shear strength reaches 24% when the void contamination index (VCI) is 20%. Ballast shear strength exhibits typical non-linear characteristics, and the relationship between fitting parameters of non-linear strength criterion and pollution degrees can be fitted by an exponential function. The fitting results can be used as the basis for strength estimation of contaminated-ballast in practical engineering. Meanwhile, geogrid can reduce the maximum dilatancy of the sample; it can also reduce the peak dilatancy angle by about 0.7°～3.7°, and achieve its maximum reinforcement effect at the VCI of 20%.

In order to explore the longitudinal and lateral resistance mechanism, contribution and influence of steel rod depth and shoulder width of steel rod reinforced sleeper, and to provide a new method for increasing resilience and stability of the long sharp slope in Sichuan−Tibet railway, a series of longitudinal and lateral resistance tests were carried out. Firstly, the overall characteristics and contribution of the longitudinal and lateral resistance of the steel rod reinforced sleeper were obtained. Then, by changing steel rod depth and shoulder width, the influence of them on the longitudinal and lateral resistance of the steel rod reinforced sleeper was explored. The results show that, compared with ordinary sleepers, the longitudinal and lateral resistance of steel rod reinforced sleeper has increased. With the shoulder with of 500 mm, shoulder height of 0 and insertion depth of 400 mm, the lateral resistance of steel rod reinforced sleeper is 53.7% higher than that of ordinary sleepers with shoulder width of 500 mm and pile height of 150 mm, and longitudinal resistance is 39.2% higher. At the same time, the contribution of sleeper bottom in lateral resistance is 8% higher compared with ordinary sleepers, with 26% higher in longitudinal resistance. The insertion depth of the steel rod exerts a great influence on the resistance. For instance, with shoulder width of 500 mm and shoulder height of 0, when insertion depth changes from 100 mm to 400 mm, the increment of longitudinal is from 5.1% to 39.2%, and that of lateral resistance is from 6.1% to 53.7%, compared with ordinary sleepers with shoulder width of 500 mm and shoulder height of 150 mm. With the change of the shoulder width, the longitudinal resistance changes slightly and the lateral resistance changes more.

In order to study the influence of hardening railway ballast bed on its own dynamic characteristics, considering the real shape of ballast particles, a simulation analysis model of hardening ballast bed was established by using discrete element method. The influences of dirty hardening and hardening degree on the dynamic response of ballast bed was analyzed, and the meso-mechanism of elastic loss and stiffness increase of ballast bed caused by hardening was discussed. The results show that the hardening will increase the vibration level of ballast particles, and the more serious the hardening is, the more obvious the influence will be, and the hardening can increase the vibration acceleration of ballast particles by 20%−30%. The compaction will strengthen the impact of the train load on the ballast bed, increase the contact force between the ballast particles, which easily causes the ballast particles to break. The working performance of the track bed at different positions was amplified and the unevenness of the ballast bed stiffness was increased by the hardening. Dirty hardening materials between ballast particles can inhibit the relative movement of ballast particles, reduce the sliding fraction of ballast particles to about 50% of the normal value, and reduce the overall macroscopic deformation of the ballast bed under train load, thus showing the characteristics of increased stiffness.

The strength degradation is an important reason for the high-speed and long-distance movement of landslides. In order to explore the influence of landslide strength attenuation on landslide movement ability, taking a high-speed Vajont landslide in Italy as an example, combined with field investigation and landslide history data and based on the shear strength attenuation theory of rock and soil mass, the discontinuous deformation analysis (DDA) method is used to study the influence of sliding band strength attenuation, sliding mass strength attenuation, and their joint action on the unique movement accumulation characteristics of the Vajont landslide. The results show that the remarkable high-speed movement and unique accumulation characteristics of the Vajont landslide are the consequences of strength degradation in both the sliding band and the sliding mass. Among them, the strength attenuation of sliding band play a dominant role in the movement velocity of the landslide. When the strength of sliding band is 15.8° and 6.9°, the maximum velocity of the monitored block is 5 m/s and 19 m/s, respectively. Meanwhile, the strength degradation of sliding mass has a significant effect on its high-speed duration and greatly improves the long-distance ability of landslide movement. When the strength of sliding mass is 40.0° and 14.0°, the maximum horizontal displacement of the monitored block is 140m and 260m, respectively. The ‘en masse’ motion character in the sliding mode can explain well the good strata sequence in the slide deposit after the landslide is fully started.

With more and more tunnel projects passing through landslide areas, research on the interaction between landslide and tunnel is particularly important. To better understand the influence of landslide with different angles of sliding zone on the stress of tunnel lining structure, taking the Nanping tunnel of Datong–Zhungeer railway as an example, laboratory model tests and numerical simulations were conducted to study the influence characteristics and variation rules of tunnel lining structure stress under different angles of sliding zone (namely 0°, 10°, 20°, 30°, 40° and 50°). The results show that the smaller the angle of sliding zone, the greater the tunnel deformation, and the greater the bending moment, shear force and earth pressure acting on the tunnel lining structure. Moreover, the maximum stress appears at the arch foot, and the forces on the left and right tunnel arch structure are asymmetric, showing a phenomenon of bias pressure. The calculation of the vertical bias stress ratio of the left and right sides of the tunnel arch structure shows that at the arch shoulder position, when the angle of sliding zone is 0, the bias stress ratio is 1.17. As the angle of sliding zone increases, the difference between the left and right stresses of the tunnel lining arch structure is getting smaller and smaller, tending to balance around the arch; but at the arch foot position, the bias stress ratio gradually increases with the increase of the angle of sliding zone. The stress difference between the left and right sides of the tunnel lining linebreak left="0.7" right="1.3"/> arch structure becomes larger and larger. For the bias tunnel, the minimum bias ratio is 1.08 and the maximum bias ratio is 1.87.

To understand the dynamic characteristics and performances of bridge-tunnel transition sections with deep buried pile-plank structures (DBPPS), dynamic field tests were performed on a transition zone including a tunnel entrance, a transition section and a abutment in the Shanghai—Kunming high-speed railway to investigate its dynamic response distributions under running trains with different train types, speeds and driving directions. A numerical model considering vehicle-track-subgrade coupled interaction was then established to study the railway line smoothness along the transition zone and the vertical dynamic stress distribution of the DBPPS subgrade. Results show that under the train loads with different train types, the maximum effective values of acceleration and displacement along the transition zone are 0.85 m/s² and 0.034 mm, respectively. The vibration level of the transition section is lower than that of the tunnel and the abutment. The effective values of dynamic response in the transition section increase with the increasing train speed, and its increase rate is smaller than that of the tunnel and the abutment. The driving directions have a significant influence on the dynamic responses in the connection between the transition section and the abutment, but have a weak influence on other sections. When the train passes through the transition zone at a speed of 300 km/h, the maximum change rate of rail deflection is approximately 0.149 mm/m, and the maximum vertical acceleration of the carbody is 0.74 m/s². The pile-plank structure can transfer the train load to the deep foundation and reduce the dynamic effect on shallow soil of the foundation.

Development characteristics and evolution law of the rock bending belts during the slope toppling deformation process is the key tostudying the geological development and mechanical behaviour of counter-tilt layered rock slopes in the process of deformation and failure. In order to reveal the influence of the slope angle change on the development law of the bending belts, the toppling deformation body in front of the dam of Gushui Hydropower Station of Lancang River is used as a prototype to generalize three sets of slope centrifugal test models with different slope angles. The centrifugal simulation tests were conducted in an effort to reproduce the evolution process of the toppling deformation of the counter-tilt layered rock slope and analyze the development characteristics of the rock bending belts inside the slope. The results indicate that the slope toppling deformation mainly occurs above the dumping and breaking datum plane, the angle between the datum plane and the normal of the rock bedding planes lies between 12° and 16°, and the value does not change with slope angle. In the process of slope deformation, the bending belts extends from the foot of the slope to the top of the slope in a stepped manner until it penetrates. A new secondary bending belts in the dumped rock mass generate above the primary bending belts in steep slope, and the slope failure mode gradually changes from single-level break to multi-level breaks. The evolution process of the bending belts can be generalized into three stages: rock layer toppling deformation, slope foot rupture-bending belts extending to the slope top, bending belts penetration to critical instability of slope. The slope anglechange has great influence on the development characteristics of slope toppling deformation. As the slope angle increases, slope toppling deformation expands and aggravates.

Affected by factors such as huge size and complex geological conditions, the super-large anchorage caisson foundation has high quality and safety risks in the construction stage. In order to assess these risks accurately, the work breakdown structure-risk breakdown structure (WBS-RBS) method, combined with expert investigation, was used to identify risks in the whole construction process. Based on the breakdown structure, an initial risk breakdown matrix was determined by estimating probabilities and corresponding loss levels of risk events. By analyzing weights of all kinds of risks through fuzzy analytical hierarchy process (FAHP), the result of the risk breakdown matrix was revised. Finally, the risk levels were divided according to the revised risk matrix values. As a case study, the proposed method was applied to construction risk assessment of the north anchorage caisson foundation of Wufengshan Yangtze River Bridge in Lianyungang−Zhenjiang high-speed railway. The results reveal that 158 risk sources were identified by the WBS-RBS and expert investigation method. Based on calculation of initial risk values and comprehensive weights, 16 major risk sources with the revised risk value larger than 1.696, such as caisson sudden sinking and attitude deviation were obtained according to the risk classification criterion. Besides, countermeasures such as special construction scheme, real-time construction monitoring, and reserved risk plan were provided. The research work provides a basis for the construction risk control of the open caisson foundation project and also provides a reference for the risk assessment and control of similar projects.

To investigate the penetration mechanism of jacked piles in laminated clay and utilize the advantages of discrete element PFC^2D software in dealing with large deformation and nonlinearity, given that the contact bonding model excels in soil simulation, a discrete element model for jacked pile penetration in laminated clay is established. The penetration process of jacked piles is realized in discrete element software. It is explored about how the pile force, pile end resistance, friction resistance and radial pressure at pile sides change with penetration depth in jacked pile penetration. The distribution characteristics of soil contact force chain are analyzed in the microscopic level for jacked pile penetration of different pile diameters. The variation law of soil displacement in jacking process is clarified. The test results show that, with the increase in pile diameter, the influence of soil layer on pile pressure decreases gradually. The friction resistance and radial earth pressure at pile sides are similar, and there is obvious degeneration at the given depth. Soil layers vary in the forms of contact force chains. When the pile end is located in the silty clay, the influence range of the pile end is about 7D (D is pile diameter); when the pile end is located in the silty soil, it is about 9D. The clay particles in silty clay mainly show radial displacement. The displacement of soil particles in the silty soil is restricted by the hardness of the upper and lower soil layers.

Formation water pressure and seepage volume in front of excavation face are important parameters of underwater tunnel projects, and also key indexes for designing the support force of excavation face and advanced water blocking and reinforcement parameters. Firstly, in view of the limitation in analyzing the 3D seepage field in front of the excavation face of underwater tunnels with the previous 2D seepage model, when building the three-dimensional seepage analytical model of the underwater tunnel, the seepage equipotential surface in front of the excavation face is considered as the space surface. Secondly, with the plane of the excavation surface as the boundary, the distribution function of water head in the half stratum space of all unexcavated areas is deduced to derive the calculation formulas of water seepage at the excavation face and pore water pressure in the front formation. Finally, the influence of factors on the water pressure and seepage in front of excavation face is analyzed, such as advanced reinforcement thickness and the relative value of permeability coefficient between soil and advanced reinforcement. The results show that, compared with the previous theoretical model of 2D seepage field in front of excavation face, the 3D analytical model can better reflect the spatial seepage pattern in front of excavation face, and the error between the calculation results of water inflow and water pressure is less than 5%; when the reinforcement range in front of excavation face is twice the tunnel diameter, and the relative permeability coefficient between the advanced reinforcement area and the stratum is set as 50, which is a safe and reasonable parameter for advanced water blocking and reinforcement.

Due to the large scale and long term of snow disaster in highway of the cold regions, the highway system in Heilongjiang Province and its frequent snowdrift disasters were taken as research objects to explore the macro conditions of disaster formation from the aspect of disaster occurrence mechanism. Based on the probability theory, the possibility, level and intensity of the hazard were linked to clarify the theory of hazard assessment through the joint probability density function of the influencing factors. After finishing the vulnerability assessment of highway transportation network, the snowdrift disaster risk assessment system suitable for cold highway network was constructed. The snowdrift events in February 2017 were evaluated, and compared the prediction results with the road warning issued by the Provincial Meteorological Information Center during this period. The results show that the highest relative risk of the road in the study area during the period is the Hegang−Yichun line, the Suihua section of the Suihua−Daqing line, the Harbin and Jiamusi sections of the Harbin−Tongjiang line, and the surrounding areas of Mudanjiang City. Their relative risk probability is more than 0.69, We need to pay more attention to this section. To a certain extent, this study provides a reference for snow disaster relief and material allocation of highway network in cold regions in the future, and makes the task of disaster prevention and mitigation in cold regions more targeted.

Two kinds of Class F fly ash-slag based geopolymer concrete, namely GPC-10 (10% slag content, 80 °C high-temperature curing) and GPC-50 (50% slag content, standard curing), are characterized by good mechanical properties. To understand better their carbonation resistance, rapid carbonation tests were carried out on the two kinds of geopolymer concrete in comparison with the ordinary Portland cement concrete (OPCC). The damage of concrete was evaluated in terms of compressive strength and splitting tensile strength. To analyze the damage cause, the composition and pore structure of the carbonized materials were investigated using X-ray energy spectroscopy (EDS) and mercury intrusion test (MIP), respectively. On this basis, carbonation models of the two kinds of concrete were established. The results show that compared with the OPCC, the geopolymer concrete has a weak carbonation resistance, especially for GPC-50, the type with high calcium content, in which the main product C—A—S—H was decomposed during carbonation, leading to an increase in porosity and thus accelerating the carbonation rate, and the carbonation depth has a linear relation with time. The 28 d carbonization depth of OPCC, GPC-10 and GPC-50 reached 2.0, 9.2 and 18.8 mm, respectively.

Straight-line fitting has received extensive attention both in curve fitting research and engineering practice. The methods of ordinary least squares and orthogonal least squares fitting ignore the existence of the observation error correlation. The coordinate pairs of surveying points, obtained by a total station in railway realignment, not only have different levels of precision but also have correlated noise. Meanwhile, straight-line fitting is usually under constraints in the realignment. Thus, a straight-line fitting model was derived based on the maximum likelihood estimation and Lagrange conditional extremum theory, considering constraints and correlated noise between coordinate components, and a Gauss-Newton algorithm was presented to search for the optimum. The method was tested with the field surveying data. Experimental results show that the proposed fitting method is capable of estimating straight-line parameters and their precisions in all circumstances by specifying stochastic models. When considering correlated noise, the precision of estimated parameters improve 9.2% with a constraint and improve 2.7% without constraints, respectively. The Gauss-Newton algorithm takes only 6 and 3 iteration times with a constraint and without constraints respectively, for locating the optimum straight-line.

To improve the ride comfort of cooperative adaptive cruise control (CACC) system under the mixed traffic flow that comprises connected automated vehicle (CAV) and manual vehicle (MV), a dual-layer control strategy considering ride comfort (RC-DCS) is proposed. From a macro perspective, the upper controller adopts a two-state space model to adjust the following distance and speed, and improve the overall stability and comfort of the fleet by the use of the cost function. From a microscopic perspective, the lower controller optimizes the logic of switching the throttle and brake pedal of a single vehicle, and stabilizes its actual acceleration output, thereby reducing the pitch caused by frequent acceleration and deceleration. The experimental results show that, the RC-DCS can reduce the following distance error and acceleration by 72.44% and 24.87% respectively in following MV condition. In the condition of MV cut-in CACC fleet, the acceleration fluctuation is reduced by increasing the following headway of 0.4 s. In the three typical conditions of vehicle following, emergency braking and cut-in, the standard deviation of the single-vehicle acceleration is reduced by 9.6%, 10.4% and 2.9%, respectively.

In order to study the parameter matching law of yaw damper for the lateral stability of high-speed trains under different wheel-rail contact conditions, a simplified model of vehicle lateral dynamics was established aiming at the typical parameters of high-speed trains operating in China. Considering the lateral stability of vehicles under the high or low wheel-rail contact conicity states respectively, the multi-objective optimization method was used to optimize the stiffness and damping parameters of yaw damper, and the influencing factors of the optimal parameters of yaw damper were analyzed as well. The results show that the optimal damping value of yaw damper mainly depends on the lateral damping of the secondary suspension, and two types of damping value selection for yaw damper are obtained. That is, when the secondary lateral damping is small, a small damping value of 600−1 000 kN•s•m⁻¹ in one side of bogie should be selected. On the contrary, the yaw damper greater than 4 000 kN•s•m⁻¹ should match the vehicle adopting a large secondary lateral damping. The stiffness of yaw damper significantly affects the stability of vehicles in different wheel-rail contact states. A smaller stiffness is conductive to the lateral stability of vehicles in low conicity wheel-rail contact state, and vice versa.

In order to understand the abnormal wear mechanism of the contact wire and the pantograph strip in the rigid catenary system, and to improve the service life of its friction elements, tests were carried out on the current-carrying friction and wear performance of the rigid catenary system with a block-on-ring current-carrying friction and wear tester. The test parameters were set as: the test current of 0, 400 A (DC), sliding speed of 60 km/h, the contact normal force of 30 N, and sliding distance of 900 km. During the tests, the sliding friction coefficient between the pantograph strip and catenary wire, the wear volume and the temperature of the pantograph strip, arc current and voltage were measured, and the wear rate of the pantograph strip and arc energy were calculated. The test results show that the tested wear rate of the pantograph strip is close to the one for actual subway lines. The DC current has a serious impact on the wear and temperature of the pantograph strip. The wear rate and temperature of the pantograph strip with DC 400 A are about 53.0 times and 7.4 times higher than those without electric current, respectively. The serious wear of the pantograph strip is mainly caused by arc ablation and adhesive wear.

In order to fully understand the service performance of ER8C and ER8 wheels for high-speed electric multiple units (EMUs), the chemical composition, the conventional mechanical properties, the fatigue properties, the impact properties, the ductile-brittle transition temperature, the fracture toughness, the fatigue crack growth threshold, the fatigue crack growth rate and etc of in-serving and over-served wheels materials were measured by electric spark direct reading spectrometer(OBLFQSN 750). The metallographic structure of the both wheels materials was analyzed,and the service performance of the both wheels were assessed. The results show as follows: 1） Compared with the ER8 wheel, the ER8C wheel has 2.74 times higher Si content, 1.29 times higher Mn content, 45% lower total Cr + Mo + Ni content and slightly lower C content; their tensile and yield strengths have increased by 5%, and their fatigue strength has increased by 15%. 2）The microstructure of the wheels contains pearlite and a small amount of ferrite, the same depth under the tread, the ferrite in the ER8C wheel is more uniform and fine, the grain size is greater than 8.5, and the ferrite wheel in the ER8 material is coarse and the grain size is grade 8. 3) The ductile-brittle transition temperature of the ER8C wheel is 84.30 ℃ and that of the ER8 wheel is 71.97 ℃. The working condition of wheels is in the brittle zone of two materials. 4) The fracture toughness of the ER8C wheel is about 6% higher than that of the ER8 wheel. However, the crack propagation threshold of the ER8 wheel is 17% higher than that of the ER8C wheel, where their crack growth rates are almost the same. 5) During service, the vertical wear rate of the ER8C wheel is slightly higher than that of the ER8 wheel.

Studying urban rail vehicles from the perspective of industrial design helps to improve train comfortability, safety, aesthetics, and cultural characteristics, so as to promote the image of the cities. Firstly, the main shapes and characteristics of six types of typical urban rail vehicles in China are illustrated with the relevant standards. Then, the key points of exterior design are summarized including the general constraints and key parameter ranges that should be considered in the form design of leading vehicles, tail vehicles and middle vehicles, and car body coating design. Next, for the interior design including driving interface and passenger interface, the existing related literatures are reviewed and the status quo and focus of the two interfaces are clarified. Finally, the research on the industrial design of urban rail trains in China is summarized and forecasted from three following aspects: multi-element coupling design theory and methodology by considering the factors of vehicles, lines, and humanity, optimization design and comprehensive evaluation technology for both interior space and passenger interface as well as interior visual environment of the vehicles.

In order to configure and evaluate the train collision energy scheme more scientifically and comprehensively, the maximum average acceleration, maximum instantaneous acceleration and energy utilization rate of trains were taken as evaluation indexes, and a more reasonable comprehensive weight coefficient was derived by using the subjective and objective index weights obtained by the order relation analysis and entropy method. A comprehensive evaluation model of train one-dimensional collision energy management was obtained by fusing weight information and evaluation index information with nonlinear weighted synthesis method, then 15 energy allocation schemes of a certain type of train were comprehensively evaluated by using this model. The results indicate that the comprehensive evaluation results are consistent with that of the change range of all indexes, and it has better applicability and reliability. Scheme 4 is the optimal scheme, in which the maximum instantaneous acceleration index decreases by 14.03%, but the maximum average acceleration index and energy utilization rate index increase by 4.92% and 19.44% compared with scheme 8 respectively. These results both reflect the importance of the maximum average acceleration index, which has the greatest impact on safety, and comprehensively considers the energy utilization rate index related to economy.

Piecewise Logistic map (PLM) is introduced into 2D coupled map lattices (2DCML) model to face a tradeoff between complexity and efficiency of chaotic systems, and a transformation method based on transient status values is used to make the output sequence of the model obey a uniform distribution. Then, the T2DCML model is built. According to the T2DCML model, an image encryption algorithm is proposed, in which encryption algorithm, the pseudo-random sequence output by the model is used to construct two elementary transformation matrices and the image is scrambled by the matrices. Then, the bits of the state value are extracted from the model to construct an integer sequence and the scrambled image is diffused by the integer sequence. Finally, the encrypted image is produced by multi-rounds of confusion and diffusion. Simulation experiments and performance analysis show that the absolute average correlation coefficient of the algorithm is 0.001 3, the information entropy is 7.999 3, and the number of pixel change rate (NPCR) and the unified average change intensity (UACI) is 99.63% and 33.60%, respectively, revealing that the algorithm can effectively meet the needs of the safe transmission of images in the network.

To improve clustering adaptability, real-time performance, and tracking accuracy in the process of dynamic obstacle detection and tracking with lidar, an adaptive density clustering algorithm and a multi-feature data association method are developed for detection and tracking respectively. Firstly, the point cloud collected by the lidar is pre-processed such as curb detection, area-of-interest extraction and ground segmentation to remove irrelevant point clouds. Then the non-ground point clouds are clustered according to the adaptive clustering algorithm to complete the obstacle point cloud detection. Finally, a weighted multi-feature data association algorithm combined with Kalman filter is used to track dynamic obstacles. The experiments show that the entire detection process with the proposed methods can accurately and stably detect and track the 10 Hz lidar data, and shorten the clustering time by 32%.

A multi-objective optimization model of siting and sizing of inverter feedback devices is established with an objective of saving the investment cost of inverter feedback devices and improving the utilization rate of regenerative braking energy. The AC-DC hybrid power flow algorithm that involves the intermittent work cycle of the inverter feedback device and the fast non-dominated sorting genetic algorithm Ⅱ (fast NSGA-Ⅱ) are combined to solve the Pareto solution set. The entropy-based technique for order preference by similarity to ideal solution (TOPSIS) is adopted to select the optimal site of the inverter feedback device. Cases with a metro line in Guangzhou were studied to compare the optimal solution with the actual configuration scheme of the inverter feedback device, showing that the optimal solution saved 700, 000 yuan of investment cost, increased the system level energy saving rate by 3.25%, and shortened the investment return period.