Abstract:In order to predict the moment-rotation (M-θ) curve of high strength steel endplate connections at ambient temperature, in fire and after fire, a calculation method for initial stiffness, post-yield stiffness, ultimate flexural resistance and shape parameter is proposed based on the four parameter exponential function, the concept of component method, and equivalent T-stubs. The prediction of moment-rotation curve is then obtained by substituting calculation results into the four parameter exponential function. Moreover, the calculation results and predicted moment-rotation curves are compared with the test results of Q690 and Q960 high strength steel endplate connections at ambient temperature, in fire and after fire. The comparisons demonstrate that the calculation values of flexural resistance are in good agreement with the test values, particularly the relative errors of Q960 are within 5%. Besides, the calculation values of initial stiffness are closer to the test results than those of EN 1993—1?8. Moreover, the predicted moment-rotation curves of Q960 are consistent with the test results and those of Q690 are relatively conservative.

Abstract:A modified modal pushover method is proposed to improve the efficiency and accuracy of seismic displacement demand calculation of cylindrical latticed shells. First, based on the energy equivalence rule, an equivalent single-degree-of-freedom (ESDF) system for spatial structures is established, in which the equivalent force and displacement of the ESDF system can be directly calculated without using interim parameters or step-by-step calculation. Then, according to the equal-energy rule, an iterative equivalence method is proposed to calculate the target displacement of the ESDF system. Finally, the overall structural seismic demand is obtained by combining the peak responses of dominant modal ESDF systems using the complete quadratic combination (CQC) rule. The numerical results show that compared to the results obtained by using the traditional pushover method, both the maximum and the envelope nodal displacement obtained by using the proposed method yield less error. The accuracy of the proposed method remains stable when structural response is dominated by multiple modes.

Abstract:The aerodynamic performance of long-span parallel twin-deck bridge is more complicated than that of single deck bridge as a result of the interference effects between two decks. By using sectional-model wind tunnel tests and two-dimensional three-degree-of-freedom (2D-3DOF) method, flutter performances and mechanisms for a long-span parallel twin-deck bridge with two commonly used sections, the composite girder and box girder, were investigated. The results show that the effects of wind attack angle on twin-deck bridges with composite girder sections are significant. Flutter performances of twin-deck bridge worsens considerably at 0 degree angle of attack, while the aerodynamic interference has little effect on those at large angle of attack. However, effects of wind attack angle on bridge with box girder section is not obvious. The flutter critical wind speeds of twin-deck bridges are 20% to 25% lower than those of single-deck bridges. Aerodynamic interference effects not only decrease flutter critical wind speeds, but also alter flutter modes. Flutter mechanisms corresponding to two section types are different, and those for upstream and downstream decks are also changed by aerodynamic interference.

Abstract:The effects of recycled coarse aggregate(RCA) replacement percentage, polymer damping material, boundary conditions and damage level on the damping properties of recycled aggregate concrete (RAC) material and slabs were studied by using suspension, cantilever, and simply-supported tests. On the material level, the test results show that the damping ratio of RAC increases with the increase of the replacement percentage of RCA. The damping ratio of RAC with 100% replacement of the RCA is increased by 30% compared with the referenced natural aggregate concrete. Adding rubber particles can significantly improve the damping ratio of RAC in the elastic stage. Adding polypropylene fiber has no remarkable effect on the damping ratio of RAC in the elastic stage. On recycled aggregate concrete slabs, it is found that the damping ratios of recycled concrete slabs increase first and then decrease with the increase of the damage degree. The relationship between the cracking behaviors and the change in damping ratios are identified by observing the development of cracks. It shows that the damping ratio of RAC slabs is related to the crack behaviors, thus the interface friction damping caused by the crack is an important factor affecting the damping ratio of RAC slabs.

Abstract:Mahavira Hall of Huishan Temple was built in the Yuan Dynasty. It is a traditional timber beam-lifting structure whose architectural form and structural technology are of great significance for study. Based on the component characteristics of the Dougong in the Mahavira Hall, six full-scale specimens were designed, two of which were tested in vertical monotonic loading while the other four were tested at cyclic loading. Failure modes and mechanical performance were studied based on the experiment and the test results. The results showed that the Dougong has a good vertical bearing capacity. However, the intersection joints of Ludou, Touang, and Nidaogong are the weak parts. In horizontal cyclic tests, failure modes were found to be the loose of the connections between components. Stiffness degradation was observed while deformation capacity, energy dissipation, and ductility were very decent.

Abstract:Structural condition assessment of drainage pipes has been a major concern for asset managers in maintaining the required performance of urban drainage systems. This paper proposed a neural network model combing extreme learning machine (ELM) and particle swarm optimization (PSO) to classify the structural condition status of urban drainage pipes. Besides, in an attempt to look for better classification performance, it used the PSO algorithm to optimize the input weight matrix and the hidden layer offset of ELM. Moreover, it validated the PSO-ELM model by using the dataset supplied from the Yangshan Bonded Port Area in Shanghai. Furthermore, it compared the predictive performance of PSO-ELM with ELM on the same dataset. The result shows that the PSO-ELM can achieve a higher classification accuracy by utilizing less neuron nodes in the hidden layer, and improve the fitting capability of ELM. The method proposed has feasibility and effectiveness for structural condition assessment of urban drainage pipes.

Abstract:Lane changing behaviors on the expressway were studied by means of the driving simulator in Tongji University. The survival analysis method was adopted to study the distribution characteristics of the time-to-line crossing (TLC) in lane changing, and establish the Cox proportional hazards model by conducting semi-parametric survival analysis, to quantitatively analyze the influence of potential factors on TLC. The results show that factors such as the adhesion coefficient of the pavements, the car-following state, and the driving speed of the vehicle have significant influences on the TLC value. The hazard rate of the vehicle driving following the other to finish line crossing increases by 1.29 times of the one driving freely. Besides, for each 0.1 increase in road friction coefficient or 1 km ? h^-1 increase in vehicle speed, the hazard rate increases by 1.14 times and 1.02 times respectively, with the TLC being shortened accordingly. However, when the initial position of lane changing is 1m further away from the target lane, the hazard rate is reduced by 0.23 times and the TLC prolonged. In addition, there is no significant difference between the TLC in different directions of lane changing.

Abstract:This paper aims to reveal the damage mechanism of urban and rural road systems in the southwestern mountainous areas in China in multi-disaster scenarios, and attempts to improve the reliable services capacity of road network in the disaster-prone areas. By selecting the Kangding section of the Dadu River Basin in the typical southwest disaster-prone zone, and by applying the complex network and computer simulation analysis method, it establishes the regional road network model and the multi-disaster interference simulation mechanism, and considers the overall connectivity and high-efficiency connectivity as the index system. The multi-disaster scenario simulates the reliability dynamic response law of road network in simulated disturbance. The empirical findings show that the reliability of urban and rural road network in the disaster-prone zone of southwestern mountainous areas is affected by the geographical location of critical road sections and geological disaster points, and the critical sections of different grades demonstrate the characteristics of cluster distribution. Based on these findings, this paper proposes some relevant countermeasures.

Abstract:The advent of the online ride-hailing platform enables the driver to obtain the departure time and destination of passengers in advance, thereby improving matching efficiency. However, it also causes potential discrimination because it makes it possible for the driver to select or reject specific passengers. In order to study the impact of information mechanism on taxi drivers, this paper uses the taxi operation data generated under different conditions by information mechanism to analyze driver service characteristics by using the descriptive statistical methods. Based on the results of non-parametric tests, taxi orders are divided into short and long distances. Spatial statistics and spatial analysis are used to study drivers’ responses to long-distance taxi orders. The results demonstrate that the impact of information differences on taxi operations. Drivers who provide the service through ride-hailing apps are likely to choose specific departure times and passenger locations when responding to long-distance taxi orders. Therefore, information regulation should be considered in the mechanism design and strategy evaluation of ride-hailing platforms in the future travel market.

Abstract:Low speed intelligent electric vehicle has been developing rapidly in recent years. The integrated localization method is the key technology in which the heading angle estimation is an important component. Based on low speed intelligent electric vehicle, a heading angle estimation method using the loosely coupled GNSS(global navigation satellite system)/IMU(inertial measurement unit) integration method is proposed in this paper. First, the heading angle estimation method is introduced. A heading angle integration method is proposed and the heading angle error dynamics and its measurement equation are derived. Aiming at the problem of time-varying quality of the GNSS signal, a residual adaptive estimation(RAE) Kalman filter is used. Then, an adaptive feedback strategy of heading error is adopted to reject the abnormal GNSS signal. Finally, under different GNSS signal conditions, the validity of the heading angle estimation method proposed in this paper is verified by multiple sets of real vehicle tests under different GNSS signal conditions.

Abstract:Experiments of diesel flame propagation and soot formation characteristics at different simulated altitudes were conducted in a constant volume combustion vessel reproducing diesel-like thermodynamic conditions of a heavy-duty diesel engine. The results show that with the increase of altitude from 0 m to 4 500 m, the ignition delay becomes longer from 0.67 ms to 1.04 ms and the ignition distance from 22.09 mm to 37.03 mm. Besides, the lift-off length increases from 23.1 mm to 34.5 mm, making the stoichiometric air ratio increase from 12.0 % to 14.0 %. Moreover, the peak value of spatially integrated natural luminosity decreases, showing the same trend of variation with the stoichiometric air ratio. Furthermore, the time integrated natural luminosity decreases, which implies that the soot formation is reduced.

Abstract:Using the CFD simulation method based on the volume of fluid (VOF) two-phase flow model, the characteristics of a multi-channel flat heat pipe developed for a lithium-ion battery was studied. Compared with the circular channel structure, the rectangular channel structure has a higher mass transfer rate of condensation/evaporation and a higher average velocity in the channels. The equivalent thermal conductivity of 6 square-channel flat heat pipe is 30.5% larger than that of the original structure with 6 circular channels. The heat pipe performance increases with the increase of channel number, but the increasing rate slows down gradually. With the fixed geometric parameters and the power input of the evaporation section, the performance of the multi-channel flat heat pipe is determined by the combined effect of condensation in the condensation section and the return efficiency of condensate liquid flow. As a result, the increase of the length of the condensation section does not substantially change the temperature difference between the two ends of the heat pipe, and the equivalent thermal conductivity does not reflect the effect of the length of condensation section on the performance of the heat pipe.

Abstract:To explore the flow field of the vehicle and the strategies of drag reduction, this paper focuses mainly on numerical simulation of the flow structures of D-shaped bluff body. Based on the theory of zero mass synthetic jet, it studies the active flow control and the active-passive flow control of D-shaped bluff body. It is found that high frequency jet can reduce the drag of D-shaped bluff body, and the maximum drag reduction can reach 1.78%. The combination of active and passive splitter plates with tabs and synthetic jet strengthens the three-dimensional flow structure of wake, impedes the extended flow structure in the wake of quasi-three-dimensional model, and makes the drag reduction perform better. The drag reduction effect is 20.86% at low frequencies and 21.20% at high frequencies.

Abstract:The torque control method of an electromagnetic eddy current brake was studied. First, the calculation formula of braking torque was obtained by using the magnetic circuit analysis method, and the mathematical relationship between braking torque and excitation current and train speed was obtained. Thus the overall scheme and operation principle of the control system were demonstrated, the main modules of the system were introduced, and the simulation model was established. Secondly, the design process and three control algorithms, which include the PID control, the fuzzy control, and the fuzzy self-adaptive PI control, were introduced. Lastly, the simulation was conducted and a comparison analysis was made between the results. The result shows that the three closed-loop control algorithms have effectively controlled the brake torque in the braking process. Moreover, fuzzy self-adaptive PI control presents the best results in both steady-state performance and transient performance, trailed by fuzzy control and PID control successively.

Abstract:To solve local vibration problems of floor in high-speed EMU(electric multiple unit), a line test is conducted in which it is found that there exists local vibration amplification of 33 Hz in the floor. Based on the principle of multi-degree-of-freedom structure vibration, the floor vibration model with multiple dynamic vibration absorbers is established. The vibration control model is used to analyze the optimal tuning parameters of the multi-dynamic vibration absorber to control the floor vibration. The control effect of the dynamic vibration absorber is verified. The results show that the installation position has a significant impact on the control effect. The vibration absorber should be installed by selecting the maximum vibration mode, and the control effect of the DVA parameter does not increase infinitely as the parameters change. The floor dynamic vibration absorber scheme can effectively reduce the vibration at the target frequency, the maximum peak value of the floor time domain is reduced by about 66%, and the comfort index is reduced by 0.3.

Abstract:Generative adversarial networks (GAN) is an excellent generative model, which can learn high-dimensional and complex real data distribution without relying on any prior assumptions. This powerful performance makes it a research hotspot in recent years, and remarkable progress has been made in research in many application fields. In this paper, the basic principle of the GAN, various objective functions and common model structures are introduced. Then, the evolutional methods for generating images under the constraints of conditional generative adversarial networks are analyzed in detail. After that, the applications of the GAN in different fields are introduced, including high-resolution image generation, small target detection, non-image data generation, medical image segmentation and so on. Finally, the optimization techniques in the training process of the GAN are summarized. The purpose of this paper is to elucidate the basic theory and development history of GAN, and to forcast the future work from the perspective of application.

Abstract:In this paper, based on finite element software Abaqus, a three-dimensional adaptive integrated modeling and simulation algorithm is established for the mirror assembly process of a conical nested Wolter-I X-ray focusing telescope. First, using Python language combined with Abaqus script interface for secondary development, an adaptive modeling simulation algorithm is proposed, and the stress distribution and surface deviation distribution of the mirror under different working conditions is accurately obtained. Then, the graphical user interface (GUI) is developed to build a more operability plug-in, which greatly improves the simulation efficiency and accuracy, and saves experimental and labor costs. The results give the ideal mirror thickness and optimal assembly load. Finally, a more reasonable evaluation method of surface deviation is obtained, which provides an important theoretical basis for optical path analysis and guidance for the actual assembly process.

Abstract:In this paper, a resource investment problem is addressed based on employee-timetabling and according to the employee timetabling in practical production systems. A mathematical model aimed at minimizing resource investment is proposed for this problem. In order to solve this problem more efficiently, the mathematical model of the original problem is simplified according to the resource investment and the properties of employee-timetabling constraints. The model proposed is proved to have the same optimal solution as the original mathematical model, and its great advantage in solving speed is verified through the CPLEX software solution process. Meanwhile, for the large-scale problem, as the constraints of employee-timetabling lead to resource occupancy between shifts, it is difficult to obtain a good solution by using the traditional activity list encoding method. A genetic algorithm with a new coding method is designed in this paper, which encodes the job delay time to search the starting time of the job. Moreover, two local optimization methods are proposed which can optimize the delay time and the starting time of jobs to improve the solution obtained by using the genetic algorithm. A comparison of the numerical experiments with CPLEX and the literature demonstrates the validity of the algorithm.

Abstract:In this paper, the pricing of a two-asset corporate bond with consideration of credit rating migration risks is studied. By using Merton’s structure approach to pricing corporate bond with default risk and giving two kinds of conditions at rating migration, two models for pricing the corporate bond are derived, which can be turned to two-dimensional parabolic equation systems coupled at the rating migration boundary. Besides, the existence and uniqueness of the solution for the model are verified. Moreover, an analytical solution and a numerical solution for the two models are obtained respectively. Furthermore, the effect of credit rating migration on pricing of corporate bond is analyzed and the difference of the two models is compared.