To minimize cross-term interferences,a method based on empirical mode decomposition(EMD) and Cohen's kernel was presented.With the method,firstly,the time-domain signal was separated into multiple intrinsic mode functions(IMFs) using EMD,then Cohen's kernel distributions of the IMFs were calculated.And a sum of all Cohen's kernel distributions was acquired.This approach was testified with three typical simulation signals and the results agreed well with the theoretical ones.At last,this approach was applied to analyze the track irregularity based on the axle-box acceleration.Compared with the results using a track inspection car,the consistent results were obtained.It was indicated that this approach is good to compensate some shortage of track inspection vehicle,and it plays an important role in high-speed high-density on-line mornitoring of track irregularity.

目前的EMD分解中,延长左右两端数据主要是通过各种数学拟合的方式（镜像延拓、多项式拟合、神经网络延拓等）来实现。在实际中,通过延长信号的采样时间,同样能够使端点数据延长,从而抑制EMD分解时的端点效应。以周期谐波函数为例,通过数值实验,将用数学拟合延长两端数据的方法（以极值点的镜像延拓为例）与直接截取两端数据的方法进行比较。同时,为了比较两种数据延长方法的效果,分别将延长后的数据进行EMD分解,将实际的EMD分解结果作为矩阵,计算与理想分解结果之间的相关系数。得出以下结论:若左右各截取半个信号周期长度的数据信号,则得到的分解结果优于通过端点延拓方法得到的EMD分解结果,且截取的点数越多,得到的结果越接近理想的分解结果。

This paper analyzed and simulated the train energy consumption algorithm based on the CBTC system simulation and performance analysis system which developed by Southwest Jiaotong University.From the simulation,the relationship of the train speed status,tracking energy and the regenerative braking energy consumption can be established on the specified civil and vehicle situation.It is necessary for the study of the energy saving strategy on this platform.

In order to solve the problem that system performances are impacted by increased train headway in the station area in an urban rail transit system,the existing algorithms for train headway in a CBTC(communication based train control) system were analyzed,and a headway optimization method was proposed based on the platform speed restriction and the speed-restriction-area parameter adjustment.With this method,an optimized train headway can be obtained by simulation of the relationship among platform speed restriction setting,train headway and travel speed.This method was simulated and tested on the CBTC simulation and performance analysis system developed by Southwest Jiaotong University.The process to optimize a train headway from 96 to 90 s in the test scenario was described.The result shows that the proposed method can optimize the headway in the critical area(s) of a CBTC system to improve the whole system performances.

In traditional way, time-frequency representation method using exponential kernel was only suitable in the condition that self-variable was along the ξ and τ axis. This paper proposed an improved time-frequency representation method to solve this problem. Using coordinates rotations, self-terms in fuzzy function could as far as possible all pass the exponential kernels function, and the cross-terms in fuzzy function could far away from the exponential kernels function. Testified this approach effected through a Chirp signal. The simulation result proved that this method can repression the interference satisfactorily, and an ideal time-frequency analysis result can be acquired.

In this paper,a novel rendering algorithm for implicit surfaces which are reconstructed from point cloud data by using a particle system was presented.First,an alternative initial technique based on bundles of parallel lines was used to find initial points that were evenly distributed on the surface.Because of its characteristics,the usual split-and-death criterion of particle system was not needed.Second,each elliptical particle was moved towards a progressively lower energy state using a conjugate gradient method,which replaced the gradient-descent as an optimization method to avoid very long convergence time and irreconcilable oscillation around the minimum.Third,a greedy selection strategy was used to choose a subset of active particles which guarantee a hole-free approximation.Finally,a relaxation process further improves the curvature driven anisotropic particle sampling.The proposed elliptical particles were specially designed for splat-based representation and could be directly converted into elliptical surface Splats without any modification,thus high-performance rendering of complex implicit surfaces can be obtained.

The energy sources hold a very large proportion in modern logistics. As a kind of clean energy source with high efficiency, the natural gas has getting more and more attention all over the world while the international oil price goes upward unceasingly and the environmental protection pressure becomes stronger and stronger. According to the energy strategy of China, 11 natural gas terminal stations will be constructed in Guangdong, Fujian, Shandong, Zhejiang, etc. to receive natural gas imported from long-distance. The large scale low temperature liquefied natural gas (LNG) tanks are specialized logistics Infrastructure; it is undoubted that their safety must be guaranteed. The liquefied natural gas tank is a complex heat transfer system. The temperature fields of large-scale low temperature liquefied natural gas tanks which independently designed by Chinese are analyzed in this paper. By using Finite Element Method, considering nonlinear heat conduction and convection, three kinds of boundary conditions and the symmetry of the tank, a model including prestressed concrete, insulating layer and LNG is built by using ANSYS software. Then, the temperature fields and temperature gradient distributions are computed at highest and lowest environmental temperatures under normal working condition. The results show that the heat preservation performance of design scheme is effective. The modeling and analysis method is helpful to other low temperature tanks.

According to the harmonious condition of deformation and the assumption of plane section,the general formula for calculating sunshine thermal stress of continuous box girders was derived.Then based on the existing experimental results and the modes of sunshine thermal gradients in the present code for design of railway and highway bridges,the practical calculating formula of thermal stress was presented.The method and formula for calculating secondary thermal bending moment were provided to analyze skew continuous box girders based on the three-moment equation of skew continuous beam.The corresponding programming was fulfilled.The thermal stresses in right and skew continuous box girders were calculated through engineering practice and good agreement was achieved compared with those of ANSYS calculation.The longitudinal distribution pattern of the thermal stress in continuous box girder bridges was ascertained.It is put forward that(1) much attention should be paid to the checking calculation for normal stress and crack resistance at midspan cross-section of central span in the design of prestrssed concrete continuous box girder bridges;(2) the checking calculation for crack resistance of inclined cross-section near gravity axis in internal support area should also be treated seriously.

在论述并行有限元技术现状和图形处理器(GPU)通用计算发展状况的基础上,通过分析对比并行有限元计算和可编程图形处理器的硬件软件结构,提出了在图形处理器上进行并行有限元计算的几个问题.文章认为,基于图形处理器的新一代开发环境CUDA完全可以实现高性能的并行有限元计算.这是一种新途径,也是一个很有潜力的发展方向.

Several severe accidents have occurred in metro transportation, the attention for safety has increased enormously. The paper introduces the method to combine metro traffic safety with Virtual Reality technology in urban rail traffic system, studies the metro fire characteristics and causes. Puts forward a method to set up the metro virtual environment for simulation by using Virtual Reality technology, which is helpful to deal with potential hazards in advance. On this foundation, by using 3DS MAX and OpenGL graphics programming techniques, the rendering pipeline of virtual metro station scene is designed. The virtual metro station scene is rendered and can be roamed and implemented in real time. The dynamical fire effect is rendered in the virtual metro station by using the real time fire simulation algorithm based on the particle system, and the locomotive and metro station fire are simulated. Preliminary fire simulation framework of the virtual metro station is established.

The characteristics and applicable situations of two common and effective error concealment methods were analyzed. An adaptive transmission error concealment algorithm for the enhancement layer in SNR scalable video coding (SVC) was proposed based on distortion estimation. The proposed algorithm estimates distortions resulted by the two concealment methods based on the correctly received reconstructed residue in the base layer and the difference between the quantization parameters of base and enhancement layers. The error concealment method resulting in a smaller distortion is selected to conceal the missing enhancement area for each 4×4 block. Simulation results of JVT SVC test model JSVM (joint scalable video model) show that average PSNR gains of 3.34 and 0.63 dB are achieved compared with the two commonly used error concealment methods, respectively.

This paper addresses the problem of reconstructing implicit function from point clouds with noise and outliers acquired with 3D scanners. We introduce a filtering operator based on mean shift scheme, which shift each point to local maximum of kernel density function, resulting in suppression of noise with different amplitudes and removal of outliers. The “clean” data points are then divided into subdomains using an adaptive octree subdivision method, and a local radial basis function is constructed at each octree leaf cell. Finally, we blend these local shape functions together with partition of unity to approximate the entire global domain. Numerical experiments demonstrate robust and high quality performance of the proposed method in processing a great variety of 3D reconstruction from point clouds containing noise and outliers.

In this paper, we present a novel rendering algorithm for implicit surfaces by using a particle system. First, an alternative initial technique based on bundles of parallel lines is used to find initial points that are evenly distributed on the surface. Because of its characteristics, the usual split-and-death criterion of particle system is not needed. Second, we move each elliptical particle towards a progressively lower energy state using a conjugate gradient method. Third, a greedy selection strategy is used to choose a subset of active particles which guarantee a hole-free approximation. Finally, a relaxation process further improves the curvature driven anisotropic particle sampling. Our elliptical particles are especially designed for splat-based representation and can be directly converted into elliptical surface splats as rendering primitives without any modification, thus we are able to obtain high-performance and high-quality rendering of complex implicit surfaces.

This paper proposes a novel fine granularity scalability coding framework with hybrid open-close loop structures (KRPS-FGS) based on key reference picture. The open-structure with single prediction loop is exploited to code non-key frames to achieve the best coding efficiency, which uses the highest quality image as the reference picture for both base and enhancement layer. To control drift, some key frames are periodically inserted, which exploit close-structure with AR-FGS to achieve the trade-off between coding efficiency and robustness. Key frames are only predicted from key frames, and no enhancement information is introduced into the prediction loop of base layer, so the drift is confined between two adjacent key frames. To optimally allocate bit-rate among enhancement layer, an optimized bit-stream extraction method is proposed, which gives more priority to the enhancement layer of key frames as compared with that of non-key frames, because the contribution of key frames to the quality of whole sequence is greater than that of non-key frames. Simulation results show that the proposed framework outperforms AR-FGS about 1~2dB over a very wide rang of bit rate. Only at the very low bit-rate points close to base layer bit-rate, the performance decreases slightly.

An algorithm for reconstructing surfaces from point clouds with noises and outliers acquired with a 3D scanner was presented.In the algorithm,each point is shifted using a filtering operator based on mean shift scheme to the local maximum of kernel density function to suppress noises and remove the outliers.The filtered data are divided into subdomains using an adaptive octree subdivision method.A local shape function is computed on each subdomain,and all the local shape functions with their weights are summed up using the method of partition of unity to approximate the global domain.Numerical experiments demonstrate that the proposed method is robust and accurate and convenient in implementing multi-resolution 3D reconstruction from point clouds data containing noises and outliers.

AR-FGS(Adaptive Reference for Fine Granular Scalable) adopts leaky prediction to achieve a better tradeoff between coding efficiency and robustness.The leaky factor is crucial to the performance of leaky prediction.This paper proposes an adaptive method for determining the optimal leaky factor for each frame.First,the disadvantage of the bit-stream extraction method used in the current JSVM(Joint Scalable Video Model) is analyzed that the bit-rate of extracted sub-steam is not smooth at the frame level.To guarantee the smoothness,a modified extraction method is presented which truncates each frame at the fixed total bit-rate.The proposed leaky factor determination algorithm sets the optimal leaky factor for each frame according to the ratio of current reference frame's base layer bit-rate to that of the first I-frame's.The optimal leaky factor is further adjusted according to the ratio of several previous frames' average base layer bit-rate to that of current reference frame's.Simulation results show that the proposed algorithm can further improve the PSNR over a wide range of bit-rate,compared with the use of the fixed leaky factor.Additionally,the bit-rate of sub-stream extracted by the modified bit-stream extraction method is smooth at the frame level.

This paper addresses the problem of reconstructing implicit function from point clouds with noise and outliers acquired with 3D scanners. We introduce a filtering operator based on mean shift scheme, which shift each point to local maximum of kernel density function, resulting in suppression of noise with different amplitudes and removal of outliers. The "clean" data points are then divided into subdomains using an adaptive octree subdivision method, and a local radial basis function is constructed at each octree leaf cell. Finally, we blend these local shape functions together with partition of unity to approximate the entire global domain. Numerical experiments demonstrate robust and high quality performance of the proposed method in processing a great variety of 3D reconstruction from point clouds containing noise and outliers.

Traditional temporal concealment techniques often simply treat the marcoblock (MB) as a whole part without any results to detailed analysis of motion characteristic of the damaged MB and any corresponding concealment strategy to this motion. That inevitably results to poor concealment performance for damaged MB, especially for those with intense and complicated motion. After the studies of the motion characteristic of the damaged MB, a novel temporal error concealment algorithm based on block warping techniques is proposed. In this algorithm, the motion of damaged MB is firstly divided into many motion cases by utilizing information of its surrounding MBs, and then different concealment methods are used to meet different motion cases. The block-warping concealment is taken once the block-warping motion of block is detected. Furthermore, the rationality of the block-warping method can be firmly ensured a by reasonable choice of the block-warping cases. Simulation results based on H.263+ codec and interlaced GOBs techniques show its satisfactory concealment performance, especially for the damaged blocks with intense and complicated motion. Compared with the popular BM and MFI-BM concealments, the subjective and objective video quality of our concealment is remarkably improved.

A novel rate model of standard quantization step size(θ) for real-time rate control is proposed based on the analysis of rate characteristic in video coding.By smoothing the θ rate curve obtained offline,a look-up table of coding rate is created first.A simple and effective rate control algorithm for real-time video is developed based on the look-up table.Because the θ rate model is an accurate model without any adaptive parameter,the θ model rate control will not suffer from the impact of scene change.Our experiments show that the performance of θ model rate control is much better than the TMN8 rate control.Comparing with the ρ domain rate control which also adopts an accurate coding rate model,the θ model rate control has lower computational complexity and offers significant advantages in hardware implementation.