Failure of high power varistor ceramic components

Ceramic varistors are electronic components which have a sharp change (over several orders of magnitude) of their electrical resistance at a well-defined voltage (switching voltage). Starting at low voltages, the resistance is large. But by exceeding the switching voltage the resistance drops dramatically and the component acts as a good conductor. Mounted parallel to a consumer, they are used as protection devices against over voltage loading. In the technical routine yearly billions of varistors are used to protect power lines, transformation stations, electronic devices, microelectronic systems or even LEDs.In service large temperature differences may come into existence, which cause severe mechanical stresses that even may destroy the components. The basic principles of the varistor behaviour are explained and examples of mechanical failure are discussed.     

基于CapsNet的汉字字形表征模型

提出基于胶囊神经网络(CapsNet)的汉字字形表征模型,通过表征汉字字形中的部件实现汉字字形的表征.首先,对任一汉字字形生成所有部件类别的表征向量.然后,根据部件存在概率,利用基于欧氏距离的离群点检测,选取相应的部件表征向量.最后,由选出的部件表征向量组成该汉字的字形表征.实验表明,文中模型在仅经过部件字形训练的情况下,即可有效识别汉字部件,同时自动生成汉字字形的有效表征.     

Parallel quality-related dynamic principal component regression method for chemical process monitoring

Traditional quality-related process monitoring mainly focuses on the magnitude change of the quality variables caused by additive faults. However, the abnormal fluctuations in the quality variables caused by multiplicative faults are often overlooked. In this paper, a novel parallel dynamic principal component regression (P-DPCR) algorithm is proposed to monitor the changes in the magnitude and fluctuation of the quality variables simultaneously. Firstly, in order to eliminate the interference of quality-unrelated variables, the quality-related process variables are selected on the basis of correlation analysis. Secondly, the dynamic extension and moving window are carried out for process variables and quality variables, in which the dynamic variables space (called X-space/Y-space) and the variance space (called VX-space/VY-space) are constructed. Afterwards, double quality-related statistics based on the regression model of these four spaces are given, and the comprehensive monitoring decision can be obtained. Finally, two numerical cases and the Tennessee Eastman process are used to show the effectiveness of the proposed method.     

结合优化支持向量机与K-means++的工控系统入侵检测方法

针对工业控制系统传统单一检测算法模型对不同攻击类型检测率和检测速度不佳的问题，提出一种优化支持向量机和K-means++算法结合的入侵检测模型。首先利用主成分分析法（PCA）对原始数据集进行预处理，消除其相关性；其次在粒子群优化（PSO）算法的基础上加入自适应变异过程避免在训练的过程中陷入局部最优解；然后利用自适应变异粒子群优化（AMPSO）算法优化支持向量机的核函数和惩罚参数；最后利用密度中心法改进K-means算法与优化后的支持向量机组合成入侵检测模型，从而实现工业控制系统的异常检测。实验结果表明，所提方法在检测速度和对各类攻击的检测率上得到明显提升。     

应用独立分量分析的激光测风雷达湍流频谱分解

针对大型风力发电机机组中常见的脉动湍流、风机尾流与涡流等湍流信号,研究了利用自然梯度下降的独立分量分析方法的湍流频谱分离效果,以区分中心风速与湍流信号,提高风机机组的综合工作效率。首先分析了风机组中常见湍流信号的后向散射与频谱分布特点,然后依据这些特点设计了对应的独立分量分析模型。在仿真结果符合要求的基础上,进行了双目激光雷达天线的风速采集与实际分离效果检测。实验结果表明,在大气折射率结构常数C2n≤10-14同时广义大气常数α≥4的通常情况下,利用双目信号能够分离出一个湍流中心和一个中心风速。对1 s内两个谱峰的波动范围进行统计,获得(2.59±0.05)MHz的中心风速以及(1.22±0.19)MHz的湍流中心估计,且二者的平均信噪比分别为25.93 dB和31.01 dB,能够在获得稳定的中心风速估计的同时得到一个较为稳定的湍流中心估计。     

A lightweight approach to component-level exception mechanism for robust android apps

Recent researches have reported that Android programs are vulnerable to unexpected exceptions. One reason is that the current design of Android platform solely depends on Java exception mechanism, which is unaware of the component-based structure of Android programs. This paper proposes a component-level exception mechanism for programmers to build robust Android programs with. With the mechanism, they can define an intra-component handler for each component to recover from exceptions, and they can propagate uncaught exceptions to caller component along the reverse of component activation flow. Theoretically, we have formalized an Android semantics with exceptions to prove the robustness property of the mechanism. In practice, we have implemented the mechanism with a domain-specific library that extends existing Android components. This lightweight approach does not demand the change of the Android platform. In our experiment with Android benchmark programs, the library is found to catch a number of runtime exceptions that would otherwise get the programs terminated abnormally. We also measure the overhead of using the library to show that it is very small. Our proposal is a new mechanism for defending Android programs from unexpected exceptions.     

休闲社区--现代居住环境景观设计手法探讨

首先提出了休闲社区的概念，提倡一种更加质朴，自然，更加清新和追求不留设计痕迹的居住环境设计风格。继而从小区地形塑造，步道系统，社区广场，植栽配置，基础绿化，水景，消防车道等各个方面，针对目前国内社区在规划设计中存在的一些弊病，对现代居住环境景观设计手法进行探讨，提出了许多具有建设性，操作性强的方法。文中还着重提到，应在建筑，道路都还没有确定的情况下，由景观规划师来做整体规划，对小区建筑，道路进行整体布局，从而创造真正的社区，营造亲切的人性空间。     

基于GIS的多源地学信息系统设计

随着地学GIS中信息的多源化,分析综合应用和评价难度也越来越大,针对GIS中的强大功能未与地学专业软件功能紧密结合,并且其中存在着多方面的问题,例如处理流程不连贯、应用方式不符合数据共享和开放GIS的思想等,在对各个软件综合分析评定的基础上,本文提出了一种符合软件工程要求一体化MGIS的系统开发与设计方案,它将适合对复杂的地学问题进行分析和处理.     

筋板类构件成形技术研究进展

筋板类构件因结构复杂,给加工带来许多困难。介绍了筋板类构件的应用概况,综述了其国内外成形技术发展现状。     

高光谱空间降采样独立成分特征分离

提出一种空间降采样独立成分特征分离方法,用于缩短独立成分分析(ICA)法在高光谱图像特征分离时的运算时间。该方法通过对高光谱图像的二维像素空间进行网格划分得到较小的窗口;基于光谱相似性测度法度量每个窗口中中心像素与周围像素的距离。然后舍弃这段距离小于阈值的周围像素,而将大于阈值的周围像素和中心像素作为样本量进行FastICA,获取投影矩阵变换原始数据,得到特征分离的ICA成分。对比了传统ICA与空间降采样(SDS)ICA(SDS_ICA)的性能,研究了降采样阈值参数、降采样窗口参数及初始投影矩阵对SDS_ICA特征分离性能及运行时间的影响。实验结果表明:应用SDS_ICA时,仅设置适中的阈值和不敏感的窗口大小参数,就能保持与传统ICA相近的特征分离性能,运行时间减少了30%以上。该方法适合应用于高光谱准实时特征提取、数据降维及目标探测等领域。