Abstract:The extended bond-based peridynamic （XPD） model resolved the limitation of fixed Poisson’s ratio in the classical model for small deformation problems. In this paper， a novel implementation using the finite deformation formulation is proposed to deal with geometrical nonlinear problems. The rotation of solid is calculated by the polar decomposition of the deformation gradient tensor derived from the least square fitting of the local displacement field. Therefore， the deformed part of the bond deflection can be separated from the general displacements， and then correct large deformation predictions can be obtained by integrating each bond force of the deformed system. Several benchmark studies are presented to demonstrate the predictive ability of the proposed model.

Abstract:The fracture of solid is still a long-standing problem in engineering and computational mechanics， which is an extremely complex phenomenon. A generalized micropolar peridynamic （GMPD） model is proposed for the three-dimensional fracture problem of quasi-brittle materials. First， the governing equations of the bonds are established by employing the Timoshenko beam theory to simulate the interaction between material points， which considers axial deformation， tangential deformation， and the relative rotational angle of bonds. Then， three kinds of peridynamic parameters， corresponding to the tensile， the shear， and the bending stiffness of the bond， are introduced to keep the consistence of the strain energy obtained from the proposed peridynamic model and from the continuum mechanics under complex loading conditions. Moreover， a novel energy-based failure criterion， involving the maximum stretch， the shear strain， and the rotation angle limits of the bond， is proposed to describe the dynamic failure for quasi-brittle materials. Finally， the three-dimensional fracture processes of the quasi-brittle materials under axial compression loads are simulated by the proposed model. The proposed model is verified by comparisons of its results and those from experimental observations， which can accurately simulate the initiation and propagation processes of different types of cracks， such as wing crack， anti-wing crack， and secondary crack.

Abstract:The stress-seepage coupling mechanism of fractured rock mass is the theoretical basis of geotechnical engineering activities such as oil and gas exploitation， in-situ stress measurement， and geological hazard prevention and control. In this paper， based on peridynamics nonlocal effects， a material points double repeat cover theoretical model was proposed. Combining the advantages of peridynamics in simulating the solid material deformation and damage and groundwater seepage， a peridynamics simulation method of fluid pressure driven stress-seepage coupling in fractured rock mass was established by using a hybrid time integration scheme and applied to the simulation of water injection test of hollow cylinder to reveal the mechanism of initiation， propagation， and connection of hydraulic fractures. The effectiveness of the simulation method was verified by comparing the results of laboratory tests with traditional numerical methods. The simulation results show that the deformation and failure of rock mass in the hollow cylinder water injection test are completely driven by hydraulic power， and the generation of hydraulic fractures is random. Therefore， there is no need to specify the fracture propagation path. In addition， the energy causing the fracture of the specimen in the process of hydraulic fracturing has a process of accumulation and release， which can be well captured by peridynamics.

Abstract:The critical issue for fluid transport simulation in non-uniform saturated porous media is to capture the mechanical behavior of fluids in the strong-weak discontinuities composed of pores， fractures， and media with different permeabilities， however， this is incompatible with the partial differential equation defined by the local fluid model in classical continuum mechanics， which leads to the difficulty of fluid simulation in saturated porous media. Based on the basic idea of unified variational peridynamics， this paper proposes a new vector state function to describe the nonlocal transport effects of fluid in saturated porous. Using the proposed fluid flux state function， this paper develops a nonlocal governing equation of fluid transport in saturated porous media. The nonlocal fluid transport model uniformly describes the mechanical behavior of fluid at the strong-weak discontinuity， avoiding the singularity of local derivative at the discontinuities. It is theoretically proved that when the radius of nonlocal effects in the nonlocal fluid model approaches to zero， the nonlocal model can be degenerated into the local ones. In order to eliminate the inherent zero energy mode problem of the nonlocal model， a fully implicit numerical scheme combined with the penalty function method is proposed to ensure the accuracy of the numerical solution. Numerical examples show that the proposed model can accurately capture the mechanical behavior of fluid passing through material interfaces， pores， and cracks. This paper provides a new model for studying the mechanical behavior of fluid transport at discontinuities in saturated porous media.

Abstract:Road profile information is used to improve vehicle suspension control and detect road irregularities such as potholes. While a great many road profile estimation approaches exist， they have been traditionally performed in a single-vehicle setting， which is inevitably susceptible to vehicle model uncertainty and measurement errors. To overcome these limitations， this paper presents a new collaborative estimation framework that exploits multiple heterogeneous vehicles to iteratively improve the estimation. Specifically， each vehicle combines its onboard measurements （e.g.， accelerometers and yaw/roll rate sensors） with a crowdsourced Gaussian process （GP） estimation from the cloud into a Kalman filter （KF） to iteratively refine the estimation. The GP is trained from the crowdsourced local estimations of all prior participating vehicles， which is then sent to the latest participating vehicle as “pseudo-measurements” to enhance the onboard estimation. The resultant local onboard estimation is sent back to the cloud to update the GP. It is shown that using the GP as additional pseudo-measurements can iteratively improve the road profile estimation performance from vehicle by vehicle. Extensive simulations are performed to show the efficacy of the proposed approach.

Abstract:Vehicular camera is an essential part of the vision system of intelligent vehicles. In harsh road or extreme conditions， due to the significant vibration of vehicles， the image sequence collected by the vehicular camera vibrates. Aimed at this problem， an electronic image stabilization algorithm for vehicle vision system is proposed. Considering the real-time requirements under vehicle conditions， the ORB algorithm （oriented fast and rotated BRIEF） is selected for feature detection and description. In order to improve the accuracy of matching and the efficiency of feature points， the traditional random sampling consistency algorithm is improved to enhance its adaptability to multiple and centralized matching points. The classical Kalman filter is sensitive to the initial value. Therefore， to adapt to the extreme conditions， the adaptive Kalman filter is used. Finally， a gasoline model vehicle with significant vibration characteristics is established， and experiments are conducted under harsh road conditions. The correctness and effectiveness of the proposed electronic image stabilization algorithm are verified under conditions more extreme than normal.

Abstract:Localization system is one of the most important parts in the environment perception system for intelligent vehicles. This paper proposed a GNSS （Global Navigation Satellites System）/IMU （Inertial Measurement Unit）/ WSS （Wheel Speed Sensor） integrated localization algorithm scheme based on tire effective radius adaption to improve the localization accuracy. First， a multi-model fusion tire effective rolling radius adaptive estimation algorithm considering wheel dynamics was designed. Then， the multi-sensor fusion integrated localization algorithm based on adaptive Kalman filter was proposed. The experimental results show that if the initial tire radius has small error， the accuracy of integrated localization algorithm can be improved by at least 30% with the tire radius adaption algorithm embedded in the system.

Abstract:Wind tunnel tests with a high frequency force balance （HFFB） for the steel tubular and angle-steel tower leg models of two typical ultra-high voltage direct-current （UHVDC） transmission towers were completed at different wind velocities. The characteristics of the aerodynamic coefficients， the skewed wind load factors， and the wind load distribution factors were studied and compared with the calculation values from different standards. The nonlinear function of the skewed wind load factor and the angle α considering the transversal lift force were constructed and the fitting parameters were also determined via the fitting analysis of the experimental values， respectively. The results show that the drag coefficient， the resultant coefficient， and the skewed wind load factor are all M-shaped distribution in skew wind. The aerodynamic coefficients of the angle-steel tower leg model are always larger than those of the steel tubular tower leg model， while the skewed wind load factors of both are generally close. Compared with the calculation values from different codes， the skewed wind load factors are underestimated at almost yaw angles. In the calculation of the wind load distribution factors in the angle-steel tower leg， the skewed wind load factor should be triangulated by using the angle α rather than the yaw angle θ. The fitting formulas proposed in this paper can be used to provide references for engineering designs.

Abstract:This paper adopts a finite element model to simulate a fire test of a double span steel portal frame and the result proves its accuracy in simulating the collapse behavior and collapse mode of double span steel portal frames. Then the finite element model verified by the fire test is used in the numerical simulation of the collapse process of double span steel portal frames. Six typical collapse modes are summarized and compared with the collapse modes of single span steel portal frames. The influence of the mid-column on the collapse mechanism of double span steel portal frames under fire is analyzed. The effects of heating conditions， stiffness of column base， connection form of mid column， load ratio， fire protection， cross-sectional temperature gradient， span-height ratio， and column spacing on the collapse mode of the frame are investigated. The results prove the accuracy of the finite element program in predicting the collapse process and collapse mode of double span steel portal frames. The parametric analysis indicates that heating conditions have a great impact on the collapse mode. Stiffness of column base， connection form of mid column， load ratio， span-height ratio， and column spacing only affect the side column collapse modes and overall collapse modes while fire protection only affects the mid-column collapse mode. The existence of cross-sectional temperature gradient rarely affects the collapse mode. This paper can provide an important basis for the early-warning of the fire-induced collapse of double span steel portal frames.

Abstract:Baihetan Hydropower Station is the largest hydropower station under construction and has the world’s largest single-unit capacity. In order to avoid the flooding of the resettlement area of Qiaojia County when the reservoir is impounded， a high filling project is required. The filling material is the locally obtained gravelly soil. In this study， the particle size analysis test， the compaction test， the compression test， and the triaxial test as well as the onsite rolling test are conducted on the gravelly soils. The results show that the compaction characteristics of the soils depend on the gravel content. When the gravel （i.e.， particle size larger than 5 mm） content is 40%~70%， the soils will form a dense skeleton structure and show better compactability and mechanical properties. During the onsite rolling compaction process， it is necessary to control the moisture content based on the gravel content of the soils and the rolling equipment adopted. Gravelly soils with compaction degrees of 0.95 and 0.97 show significant differences in their compression properties and shear strength. Meanwhile， according to the stress-strain relationships of the soils， the parameters of the Duncan-Chang hyperbolic model are determined. The test results provide an important reference for the gradation and compaction design of the filler as well as the parameters of the filler for the Baihetan high fill project.

Abstract:The suction bucket foundation of wind power is subjected to horizontal loads from wind， waves， ocean currents， earthquakes， and ship docking in service conditions， leading to failure modes of translation and rotation. The accumulated ground deformation results in severe structural instability and damages. The suction bucket foundation demonstrates a shallow embedded depth， and the soil-structure interactions are different from the traditional shallow foundations. Therefore， the failure mode of the wide-shallow foundation is not clearly investigated. A series of centrifuge tests are conducted to study the ultimate lateral bearing behavior of the wide-shallow suction bucket foundation with variable L/D ratios in sandy soil. The progressive failure mechanisms of the suction bucket foundation are further illustrated through Abaqus numerical simulations. The evolution of the rotation center is depicted. An analytical calculation method is proposed by considering the rotation center， aiming to provide design references for practical engineering.

Abstract:The amplification of discharge caused by flash flood disasters has affected the performance of Bayesian runoff simulation in hilly watersheds. This paper， taking the Baisha River Basin in Dujiangyan with frequent flash flood disasters after the Wenchuan earthquake as the research area， explores the critical rainfall causing flash flood disaster in the basin， introduces the linear amplification factor to correct the amplification effect of flash flood disaster on flood flow， and constructs six Bayesian flood simulation schemes by coupling the first-order autoregressive residual error model with different heteroscedastic transformation functions. The outlet daily runoff process of the basin is then simulated using the GR4J model. The results show that the critical rainfall and linear amplification factor can effectively identify and correct the impact of flash flood disasters on discharge. After considering the influence of flash flood disasters， the accuracy， reliability， and precision metrics of the daily runoff simulation results in the Baisha River Basin are increased by 19%， 32%， and 62%， respectively. This method can be widely used in the basin with frequent flash flood disasters and provide a more accurate scientific basis for runoff forecasting in hilly watersheds.

Abstract:Based on the beach profile survey data after the beach nourishment project on the coast between the Daihe Estuary and the Yanghe Estuary， the sandy beach under the low-energy condition with a single offshore sandbar was selected to investigate its morphological evolution process. A new equilibrium profile function with a single bar， which is comprised of background profile and barred profile， is derived. The background profile is based on the principle of maximum entropy to consider the difference between onshore and offshore side of the beach profiles， while the barred profile is described by 3 to 4 parameters to show the asymmetric sandbar shape in the form of parabolic function. The new shape function fits well with the measured post-nourishment beach profile from the Daihe Estuary to the Yanghe Estuary， which also has been successfully applied to the Tiger-Rock Beach in Qinhuangdao and the Fort Myers Beach in Florida. It indicates that the new shape function can accurately reflect the characteristic of the low-energy beach with a single sandbar.

Abstract:In order to solve the problem that the accuracy and speed of the existing methods are not high in recognition of pavement distress image， a recognition method of pavement distress image based on historical information is proposed by using historical detection results. First， an algorithm framework for the application of historical information is built， and the mechanism of using historical information to create initial constraints for distress detection is introduced. Next， the VGG-16 network is trained to extract damage features. Finally， the initial population is established， and the feature parameters are designed by using historical information to improve the genetic algorithm. The experimental results show that this method can significantly improve the recognition speed without reducing the recognition accuracy compared with the recognition method without using historical information， and the fastest speed can be 141.71 times higher than the algorithm without historical information.

Abstract:This paper aimed to study the influence of distance， density， pedestrian flow， and visibility on the heterogeneity of passenger decision-making preference in emergency evacuation in subway stations. The conditional Logit model and random parameter Logit model were used to quantify the utility coefficients of the four influencing factors based on the data collected from 20 emergency evacuation scenarios， and the heterogeneity of pedestrian decision preference was quantitatively analyzed according to the marginal probability distribution of the utility coefficients. The results show that the distance， density， and pedestrian flow have negative influences while visibility has a positive utility. The goodness of fit of the random parameter Logit model is higher than that of the conditional Logit model. The coefficients of the four influencing factors are random variables. Distance has the lowest level of heterogeneity while density and pedestrian flow have a slightly higher level of heterogeneity， and visibility has the highest level of heterogeneity.

Abstract:While granular computing， in particular fuzzy sets， offers a comprehensive conceptual framework of managing conceptual entities-information granules with the processing mechanisms provided by fuzzy sets， this mechanism does not fully engage experimental data whose characteristics could be fully incorporated in the obtained results. Having this in mind， the fundamental constructs of relational computing and fuzzy relational computing were augmented by mechanisms originating from the area of granular computing. It is advocated that the prescriptive constructs existing in the area of relational calculus are augmented by incorporating a descriptive component， which is governed by the existing data. It is shown that the principle of justifiable granularity helps enhance the existing results by experimental evidence residing with the data and develop the results in the form of information granules.

Abstract:Existing news recommendation models generally consist of the text feature extraction network and the recommendation network. News-related side information， such as category， is not fused into the text feature extraction network. Without fusing it， there are differences between the optimization targets of the text feature extraction network and the recommendation network. In this paper， a general SIACNN （side information aggregated CNN） layer is proposed. The SIACNN layer fuses the side information into the text feature through the attention mechanism， which bridges the gap between text feature extraction and recommendation tasks and improves the effectiveness of the recommendation. CNNs are replaced in many state-of-the-art models which used CNNs to extract the text feature with the SIACNN and several experiments are conducted in a large real-world news recommendation dataset MIND（MIcrosoft News Dataset） collected from MSN（MicroSoft News）. The recommendation effectiveness and generality of SIACNN are verified by several experiments.