An Ensemble Kalman Filter Approach Based on Level Set Parameterization for Acoustic Source Identification Using Multiple Frequency Information

In this paper, a reconstruction problem of the spatial dependent acoustic source from multiple frequency data is discussed. Suppose that the source function is supported on a bounded domain and the piecewise constant intensities of the source are known on the support. We characterize unknown domain by the level set technique. And the level set function can be modeled by a Hamilton-Jacobi system. We use the ensemble Kalman filter approach to analyze the system state. This method can avoid to deal with the nonlinearity directly and reduce the computation complexity. In addition, the algorithm can achieve the stable state quickly with the Hamilton-Jacobi system. From some numerical examples, we show these advantages and verify the feasibility and effectiveness.     

单宁酸/两性离子改性油-水分离膜的制备及性能

以不锈钢网为基材, 利用单宁酸对不锈钢网进行表面预处理并功能化接枝两性离子基团, 制备了新型亲水和水下超疏油的单宁酸/两性离子改性油-水分离膜(TA-ZW-SSM). 利用X射线光电子能谱仪(XPS)、 扫描电子显微镜(SEM)及接触角测量仪等表征了其化学结构、 形态和润湿性. 研究结果表明, 两性离子基团通过化学键接枝在单宁酸预处理的不锈钢网表面. 油-水分离实验结果表明, 对于不同类型的油-水混合物, 本文制备的超亲水和水下超疏油特性的TA-ZW-SSM可实现重力驱动的高效油-水分离, 并具有较好的化学稳定性及再循环性.     

基于多尺度学习与深度卷积神经网络的遥感图像土地利用分类

土地利用信息是国土资源管理的基础和重要依据，随着高分辨率遥感图像数据的日益增多，迫切需要快速准确的土地利用分类方法。目前应用较广的面向对象的分类方法对空间特征的利用尚不够充分，在特征选择上存在一定的局限性。为此，提出一种基于多尺度学习与深度卷积神经网络（deep convolutional neural network，DCNN）的多尺度神经网络（multi-scale neural network，MSNet）模型，基于残差网络构建了100层编码网络，通过并行输入实现输入图像的多尺度学习，利用膨胀卷积实现特征图像的多尺度学习，设计了一种端到端的分类网络。以浙江省0.5 m分辨率的光学航空遥感图像为数据源进行了实验，总体分类精度达91.97%，并将其与传统全卷积网络（fully convolutional networks，FCN）方法和基于支持向量机（support vector machine，SVM）的面向对象方法进行了对比，结果表明，本文所提方法分类精度更高，分类结果整体性更强。     

医学影像处理的深度学习可解释性研究进展

随着医学影像数据的迅速增长，传统的影像分析方法给医生带来巨大挑战。利用计算机视觉技术提供自动或半自动辅助诊断，可大大缓解人工阅片压力，提高诊断的准确性，促进医疗流程的标准化建设等。目前，深度学习卷积神经网络在医学影像处理中已取得不俗表现，但深度学习“黑匣子”的不可解释性阻碍了智能医疗诊断的发展。为增强对医学影像数据处理的深度学习可解释性的了解，对近几年相关研究进展进行了综述。首先，综述了深度学习在医学领域的应用现状及面临的问题，对神经网络的可解释性内涵进行了讨论；然后，从现有深度学习可解释性的常见方法出发，重点讨论了医学影像处理的深度学习可解释性研究进展；最后，探讨了医学影像处理的深度学习可解释性的发展趋势。     

An experimental method for eliminating effect of rigid out-of-plane motion on 2D-DIC

The out-of-plane motion is one of the most important factors that affect the precision of two-dimensional digital image correlation (2D-DIC). In this paper, a novel solution is presented to improve conventional 2D-DIC by eliminating the effect of out-of-plane motion, including translation and rotation. Firstly, an experimental technique using two projected laser strips is proposed to measure the out-of-plane motion of a planar specimen. A theoretical model is then established to predict the pseudostrains caused by out-of-plane motion based on the pin-hole imaging model. Using the measured out-of-plane displacement, the captured deformed images used in 2D-DIC are amended to eliminate the effect of out-of-plane motion by the theoretical model. Finally, two experiments were conducted to validate the effectiveness of the proposed method. Results indicate that application of the proposed method can effectively eliminate the errors caused by out-of-plane motion.     

Surface integral analogy approaches for predicting noise from 3D high-lift low-noise wings

Three surface integral approaches of the acoustic analogies are studied to predict the noise from three concep- tual configurations of three-dimensional high-lift low-noise wings. The approaches refer to the Kirchhoff method, the Ffowcs Williams and Hawkings （FW-H） method of the permeable integral surface and the Curle method that is known as a special case of the FW-H method. The first two approaches are used to compute the noise generated by the core flow region where the energetic structures exist. The last approach is adopted to predict the noise specially from the pressure perturbation on the wall. A new way to con- struct the integral surface that encloses the core region is proposed for the first two methods. Considering the local properties of the flow around the complex object-the actual wing with high-lift devices-the integral surface based on the vorticity is constructed to follow the flow structures. The surface location is discussed for the Kirchhoff method and the FW-H method because a common surface is used for them. The noise from the core flow region is studied on the basis of the dependent integral quantities, which are indicated by the Kirchhoff formulation and by the FW-H formulation. The role of each wall component on noise contribution is analyzed using the Curle formulation. Effects of the volume integral terms of Lighthill＇s stress tensors on the noise pre-diction are then evaluated by comparing the results of the Curle method with the other two methods.     

General decay of solutions of a wave equation with memory term and acoustic boundary condition

In this paper, the global solvability to the mixed problem involving the wave equation with memory term and acoustic boundary conditions for non‐locally reacting boundary is considered. Moreover, the general decay of the energy functionality is established by the techniques of Messaoudi. Copyright © 2016 John Wiley & Sons, Ltd.     

Pseudo‐spherical normal Darboux images of curves on a lightlike surface

We introduce the notions of the pseudospherical normal Darboux images for the curve on a lightlike surface in Minkowski 3‐space and study these Darboux images by using technics of the singularity theory. Furthermore, we give a relation between these Darboux images and Darboux frame from the viewpoint of Legendrian dualities.     

不同类型专利对经济增长影响的实证研究

文章考察三种不同类型专利对经济增长的影响,结合广东省1985-2014年最新数据,通过Matlab、Excel软件对数据进行图像模拟,并根据函数理论、计量经济学理论,创造性地构建一种多元非线性计量模型。进一步地,运用所得模型进行实证分析,结果显示发明专利在经济增长中产生的影响远大于实用新型专利和外观设计专利,论证了原创性创新在现阶段广东省经济转型升级中的核心作用。