3D Anisotropic Diffusion on GPUs by Closed-Form Local Tensor Computations

We present an efficient implementation of volumetric anisotropic image diffusion filters on modern programmable graphics processing units (GPUs), where the mathematics behind volumetric diffusion is effectively reduced to the diffusion in 2D images. We hereby avoid the computational bottleneck of a time consuming eigenvalue decomposition in $\mathbb{R}^3$. Instead, we use a projection of the Hessian matrix along the surface normal onto the tangent plane of the local isodensity surface and solve for the remaining two tangent space eigenvectors. We derive closed formulas to achieve this and prevent the GPU code from branching. We show that our most complex volumetric anisotropic diffusion filters gain a speed up of more than 600 compared to a CPU solution.     

基于小波滤波的低场NMR纤维上油率检测新方法

针对目前低场核磁共振纤维上油率检测存在的问题,提出了一种基于小波变换的纤维上油率检测新方法. 首先,用低场核磁共振采集标准纤维自由感应衰减信号进行线性拟合得到定标曲线;然后,用测试纤维采集多组叠加次数不同的核磁共振信号,并分别进行小波软阈值滤波;最后,将滤波后的多组结果代入定标曲线计算纤维的上油率. 检测结果和实际应用表明小波软阈值滤波可有效去除纤维FID信号中的噪声并大幅缩短检测时间. 此方法可为低场核磁共振纤维上油率快速、准确检测奠定基础.     

舰船单轴旋转激光捷联惯导系统动态初始对准

初始对准的时间和精度是舰船惯导系统的重要指标。针对在不同情况下惯导系统启动的实际工程需求,提出了单轴旋转激光捷联惯导系统的初始对准方案。研究了系统在动基座情况下进行粗对准方法。建立了单轴旋转惯导系统的误差模型,使用卡尔曼滤波的方法实现了系统精对准过程。分别对惯导系统三种不同动态启动条件,设计了不同的对准方案。数字仿真结果表明,经过6h的初始对准,垂向陀螺常值漂移的对准误差在设定值的5.2%以内,垂向加表零偏的对准误差在设定值的1.8%以内。     

Information filtering in evolving online networks

Recommender systems use the records of users' activities and profiles of both users and products to predict users' preferences in the future. Considerable works towards recommendation algorithms have been published to solve the problems such as accuracy, diversity, congestion, cold-start, novelty, coverage and so on. However, most of these research did not consider the temporal effects of the information included in the users' historical data. For example, the segmentation of the training set and test set was completely random, which was entirely different from the real scenario in recommender systems. More seriously, all the objects are treated as the same, regardless of the new, the popular or obsoleted products, so do the users. These data processing methods always lose useful information and mislead the understanding of the system's state. In this paper, we detailed analyzed the difference of the network structure between the traditional random division method and the temporal division method on two benchmark data sets, Netflix and MovieLens. Then three classical recommendation algorithms, Global Ranking method, Collaborative Filtering and Mass Diffusion method, were employed. The results show that all these algorithms became worse in all four key indicators, ranking score, precision, popularity and diversity, in the temporal scenario. Finally, we design a new recommendation algorithm based on both users' and objects' first appearance time in the system. Experimental results showed that the new algorithm can greatly improve the accuracy and other metrics.     

一种基于级联长周期光纤光栅的力传感装置

利用光纤布拉格光栅(FBG)具有谱线窄、反射率高的特性和级联长周期光纤光栅(级联LPFG)的边沿滤波特性, 设计了一种基于级联LPFG的力传感装置, 将拉力变化转变为FBG反射谱功率的变化. 在理论上分析了实验的可行性, 用光纤光栅刻写系统制作了符合要求的级联LPFG和FBG, 用光谱分析仪对FBG的反射谱进行了监测. 用光纤功率计对FBG的反射功率进行了测量, 并对测得的数据进行了线性拟合, 其线性相关系数为0.9924, 装置的灵敏度为1.074μW?N-1. 在拉力为0~0.6N的范围内功率与拉力有较好的线性关系. 测量范围和灵敏度取决于级联LPFG透射谱边沿的斜率. 该装置结构简单, 操作方便.     

Impact of subgrid fluid turbulence on inertial particles subject to gravity

Two-phase turbulent flows with the dispersed phase in the form of small, spherical particles are increasingly often computed with the large-eddy simulation (LES) of the carrier fluid phase, coupled to the Lagrangian tracking of particles. To enable further model development for LES with inertial particles subject to gravity, we consider direct numerical simulations of homogeneous isotropic turbulence with a large-scale forcing. Simulation results, both without filtering and in the a priori LES setting, are reported and discussed. A full (i.e. a posteriori) LES is also performed with the spectral eddy viscosity. Effects of gravity on the dispersed phase include changes in the average settling velocity due to preferential sweeping, impact on the radial distribution function and radial relative velocity, as well as direction-dependent modification of the particle velocity variance. The filtering of the fluid velocity, performed in spectral space, is shown to have a non-trivial impact on these quantities.     

受污染混沌信号的协同滤波降噪

根据混沌吸引子的自相似分形特性,提出了一种利用协同滤波重构受污染混沌信号的降噪算法.所设计的降噪算法通过对相似片段的分组将一维混沌信号的降噪转化为一个二维联合滤波问题;然后,在二维变换域用阈值法衰减噪声;最后,通过反变换获得原始信号的估计.由于分组中的相似片段具有良好的相关性,与直接在一维变换域做阈值降噪相比,分组的二维变换能获得原信号更稀疏的表示,更好地抑制噪声.仿真结果表明,该算法对原始混沌信号的重构精度和信噪比的提升都优于小波阈值、局部曲线拟合等现有的混沌信号降噪方法,对相图的还原质量也更好.     

车载紧耦合MIMUs/GPS的神经网络辅助强跟踪滤波方法

微惯性测量单元由三轴正交的微机械陀螺、加速度计和微型地磁传感器组成.将上述装置与GPS接收机组合,可构成最佳导航定位模型,其中紧耦合MIMUs/GPS对全导航参数(位置、速度及姿态)的测量精度可大幅提高.由于微惯性传感器具有大漂移特性,为获得具有自适应的线性参数模型,提出了融合滤波的信息处理方法,利用强跟踪滤波实现状态预测,二阶EKF实现测量更新,并借用神经网络技术完成对状态预测的修正.由于系统组件具有非线性,该神经网络辅助的强跟踪滤波方法旨在逼近MIMUs/GPS的真实特性,并为车载用户提供更为精准的导航参数信息.动态环境下的仿真试验表明,尽管MEMS惯性传感器的精度有限,所提出的方法能够有效用于实际的导航参数解算.     

一种新的改进Canny图像边缘检测算法

针对传统的Canny算法在平滑滤波时对图像边缘检测的影响,以及需要预先设定高低阈值等缺点,本文提出了一种新的Canny边缘检测算法。首先采用自适应中值滤波对图像进行处理;其次,添加45°和135°方向的梯度模板来计算梯度幅值;最后,采用Otsu算法计算经梯度幅值运算得到图像的高低阈值。实验结果表明,该改进的Canny算法具有良好的抗噪性能,并且能精确的检测出边缘信息。     

GNSS辅助捷联惯导行进间对准自适应滤波方法

针对车载武器系统快速发射需求,提出一种基于GNSS辅助的捷联惯导行进间对准自适应滤波方法。该方法把行进间传递对准分为粗对准与精对准两个阶段。粗对准阶段以GNSS为观测基准完成对捷联惯导姿态的粗捕获,降低初始偏差不确定性对于精对准阶段的影响。在精对准阶段,考虑到车载系统的运动特性,提出一种"水平+方位"行进间对准双滤波器并行的设计思路,利用车载系统在不同时间段的动力学特性,对三轴姿态估计进行分时解耦,实现初始姿态的高精度估计;与此同时,引入协方差成形自适应调节过程,以最小化Frobenius范数为优化指标,实现对行进间对准卡尔曼滤波器的自适应调节,增强系统鲁棒性。数值仿真表明,协方差成形自适应卡尔曼滤波方法能够有效保证系统在全运动剖面内的稳定,结合双滤波器并行方案能够有效解决行进间对准精度不高与稳定性欠佳等问题,水平对准精度优于1.5′(1σ),方位对准精度优于6′(1σ)。