集成多策略改进FCM算法的旋转机械故障数据聚类分析研究

针对旋转机械故障数据聚类分析中的初始聚类中心不确定和孤立点敏感问题,提出了一种集成多策略改进的模糊C均值(FCM)聚类方法。首先以故障数据集的决策属性为等价关系对数据集进行划分,得到若干个由等价关系导出的等价类;然后以每个等价类为可行域,采用均值漂移方法搜索故障数据类中心;最后以搜索到的类中心为FCM算法的初始聚类中心,通过核技术计算故障数据样本与相应类中心在高维特征空间中的欧氏距离,从而实现数据样本相似性的有效度量,并完成故障数据的模糊聚类。通过标准数据集和旋转机械故障数据集对方法的性能进行了验证及比较分析。结果显示,改进FCM算法的聚类性能相比传统FCM算法的聚类性能得到了明显提升,在收敛速度和聚类准确性两个性能指标上,改进的FCM算法比FCM算法具有显著优势。     

基于EEMD模糊熵和GK聚类的信号特征提取方法及应用

提出了一种基于集合经验模态分解模糊熵和GK聚类相结合的方法,应用于滚动轴承的故障诊断中。首先,利用EEMD方法将故障信号分解成多个本征模态分量来消除模态混叠影响;其次,通过相关性对IMF分量进行筛选,并求取其模糊熵作为特征向量进行GK聚类分析进行模式识别。在实验分析中,通过模糊熵、样本熵、近似熵3种特征参数的对比,和GK聚类与FCM聚类的对比,证明了该方法的有效性和优越性。     

基于EEMD样本熵的直流牵引网故障电流识别

针对地铁直流牵引网的振荡电流容易引起继电保护系统频繁误动的问题,提出一种结合总体平均经验模态分解(ensemble empirical mode decomposition,EEMD)和样本熵的直流牵引网振荡电流与短路故障电流识别方法。利用EEMD方法对直流牵引网的馈线电流信号进行分解,求取各固有模态函数(intrinsic mode function,IMF)分量的样本熵值,并将计算结果求和,进而获得反映直流牵引网运行状态信息的特征量。通过对典型馈线电流信号进行分析计算,可知EEMD和样本熵相结合的特征提取方法可以有效地区分直流牵引网振荡电流与短路故障电流。算例分析验证该方法的有效性。     

一种基于EEMD-SVD和FCM的轴承故障诊断方法

提出了一种基于总体平均经验模式分解(EEMD)和奇异值分解(SVD)的模糊C均值聚类(FCM)相结合的轴承故障诊断方法。首先对轴承信号进行EEMD分解,得到若干个平稳的本征模函数(IMF),再通过相关性分析筛选包含主要信息的前几个分量进行奇异值分解,然后将得到的奇异值矩阵作为特征向量,通过FCM模糊聚类进行识别。实验结果表明,此方法可有效地对轴承故障类型进行识别。     

石墨烯/纳米减反结构复合透明导电薄膜的研究

石墨烯具有较高的透过率及良好的电导率, 作为透明导电薄膜具有潜在的应用价值。首先在石英基底上引入SiO₂纳米球阵列结构作为光学减反射层, 使石英基底可见光区光学透过率从93.2%增加到99.0%。然后利用常压化学气相沉积方法, 通过基底表面铜颗粒远程催化碳源, 直接在减反层上可控制备具有石墨烯/纳米减反结构的新型复合透明导电薄膜。通过去除SiO₂纳米球阵列结构形成反相复制的石墨烯空心球阵列结构, 且生长时间10 min时, 对应半高宽约40 cm^-1, I_2D/I_G = 2.31, I_D/I_G = 0.77, 证明在SiO₂纳米球阵列减反结构上制备了低缺陷且连续的全包覆少层石墨烯薄膜。引入SiO₂纳米球阵列减反结构后, 其在可见光区光学550 nm波长处的透过率平均提高了5.5%, 方块电阻相对无减反射层基底平均降低了20.09%。本研究方法避免了复杂的转移工序, 减少了对石墨烯的损失与破坏, 同时实现了高透光性及高导电性的功能协同, 在光伏器件、平板显示等领域展示出更大的应用前景。     

分析城市轨道交通供电直流侧短路故障定位方法

城市轨道交通中较多采用的是低压直流上下行接触网全并联的双端供电方式向机车提供电能,由于受到系统容量和供电电压的影响,其直流侧短路电流会对设备造成较大的危害,如何快速有效的对故障位置进行确定并排除,成为了当前城市轨道交通中研究的重点。     

基于小波包变换和随机森林算法的光伏系统故障分类

针对光伏系统故障分类问题,提出一种小波包变换和随机森林算法相结合的故障分类方法。采集光伏系统的故障电压数据,利用小波包变换对电压信号进行分解,提取各频带能量作为故障特征,将特征样本送入随机森林算法中进行分类。随机森林算法是结合集成学习理论和随机子空间方法的一种算法,可以对多种故障做出准确分类。使用PSCAD/EMTDC搭建独立光伏发电系统,选取12种故障进行模拟,得到600个故障样本,选取其中360个样本用于训练分类器,240个样本用于测试分类器的分类性能。仿真结果表明:该方法可有效辨别光伏系统的12种故障,分类准确率达到97.92%。与RBF神经网络分类器相比,故障分类准确率提高了4.17%,对进一步实现光伏系统故障诊断研究具有重要意义。     

衬底温度和生长氧压对La0.67Sr0.33MnO3:Ag0.08薄膜激光感生电压效应的影响

通过化学共沉淀法制备了La_0.67Sr_0.33MnO₃:Ag_0.08 (LSMO:Ag_0.08)多晶材料, 然后采用脉冲激光沉积(PLD)技术在LaAlO₃ (LAO)倾斜衬底上制备了LSMO:Ag_0.08薄膜。研究了衬底温度和生长氧压对薄膜结构、电输运特性及激光感生电压(LIV)效应的影响。结果表明: 当衬底温度为790℃、生长氧压为45 Pa时, 薄膜具有最大峰值电压(U_p)、优值(F_m)和各向异性Seebeck系数(ΔS); 在优化的衬底温度和生长氧压条件下, 长程Jahn-Teller协变引起ΔS数值提高, 这是LIV信号增强的主要原因。     

基于遗传模糊C-均值聚类算法的除湿机故障诊断

针对除湿机系统的故障诊断问题及其特点,以CFTZ21型除湿机为对象,应用模糊C-均值聚类（FCM）算法进行了研究;引入遗传算法对传统模糊C-均值聚类算法进行了改进,克服了传统算法的不足;结合实验采集到的数据样本,对改进后的遗传模糊C-均值聚类算法进行检验,结果达到预期效果,由此说明,将改进的FCM应用于除湿机故障诊断是可行的。     

基于VMD_MPE和FCM聚类的变转速工况下转子不平衡故障诊断方法EI北大核心CSCD

旋转机械在变转速工况下转子不平衡故障诊断问题一直是故障诊断领域的难点,为解决该问题,提出一种基于变分模态分解(variational mode decomposition,VMD)与多尺度排列熵(multiscale permutation entropy,MPE)和模糊C均值(fuzzy C means,FCM)聚类结合的故障诊断方法(VMD_MPE-FCM)。首先,用VMD对转子的振动信号进行处理,得到若干本征模态分量(intrinsic mode function,IMF);然后,根据转子不平衡故障时一倍频(1×)处振幅剧烈增加的现象,从VMD得到的各IMF频谱图中筛选出最能表征转子不平衡故障特征的IMF;进而采用MPE法对筛选出的IMF进行量化;最后,将量化所得值作为特征向量输入FCM,得到各转速工况下的标准聚类中心,采用择近原则,运用模糊贴近算法计算出待识别数据与标准聚类中心的贴近度,从而实现变转速工况下转子不平衡的故障识别。在转子试验台上采用VMD_MPE-FCM法进行了变转速工况下转子不平衡故障诊断试验,试验结果表明:该方法是有效的,可以准确、高效地提取出转子故障特征,能够很好地识别出不同转速工况下转子的不平衡故障。     

Sparse convolutional autoencoder-based fault location for drive circuits in nuclear reactors

Drive circuit is a critical part of instrumentation and control systems in nuclear reactors, and its performance directly influences the operation of nuclear reactors. However, comparing with the open circuit IGBT faults, soft faults caused by the degradation of electronic components present much slighter fluctuations to the performance of drive circuits. If the two fault modes co-exist, traditional fault diagnosis models are prone to misclassify soft faults as the normal condition. To improve the accuracy of fault diagnosis of drive circuits, it necessitates to accurately locate the faults of drive circuits, while effectively extracting the distinguishable fault features is one of the critical factors for fault location. In this article, a fault location method combining the empirical modal decomposition (EMD) algorithm and sparse convolutional autoencoder (SCAE) is proposed. The EMD algorithm is applied to decompose the three-phase current signals of drive circuits. An SCAE-based feature extractor is constructed to capture high-dimensional and sparse fault feature data with the aid of the powerful feature autonomic extraction capability of deep learning. A deep classifier is designed to locate faults in the driver circuit. A fault simulation model of the drive circuit is developed and the monitor data is collected. The effectiveness of the proposed method is validated via a real case of drive circuit in nuclear reactors.     

基于小波包特征提取的模糊诊断网络建立及应用

本文首先根据小波包分解原理和应用经验总结出小波包特征量与汽机故障对照表,将其与模糊综合评判和BP网络有机结合在一起,建立了基于小波包特征提取的模糊BP诊断网络模型.采用模糊综合评判技术,使该网络可在少量典型故障样本监督下训练成功,对于缺少机组运行故障知识库的厂家具有推广应用前景.最后举例说明,该网络在汽机诊断中是一种有效的智能分类器.     

Studies on electrochemical photovoltaic cells formed with as-deposited in-doped CdS films

Indium doped and undoped CdS films have been deposited on conducting substrates by a chemical bath deposition technique. The doping concentration was varied from 0.005 to 1.0 wt. % of indium. Electrochemical photovoltaic (ECPV) cells are formed with these films and their electrical properties are studied. It is observed that the performance of the cell measured in terms of open circuit voltage (Voc), short circuit current (Isc), efficiency (η), and maximum power output (Pm), is optimum for the cells formed with 0.01 wt. % In-doped CdS photoanode.     

太阳电池填充因子随日照强度变化的理论分析与计算

根据太阳电池直流模型和最大功率点数学条件,推出短路电流、开路电压、最大功率点电流和电压 以及填充因子随日照强度变化的数学关系式;并选取２个电池分别计算出在不同日照强度下上述电池参数随日 照强度的变化率。验证了短路电流和最大功率点电流是近似跟日照强度成正比,开路电压和最大功率点电压是 近似跟日照强度的自然对数成正比。并提出了填充因子随日照强度的变化关系不具有简单的函数形式,而对不 同的太阳电池其变化关系也迥异。最后用Multisim的模拟结果检验了理论分析计算的正确性。     

Photovoltaic fiber

The optoelectronically active optical fiber is demonstrated in this work. This fiber consists of dye sensitized solar cell (DSC) structure deposited on claddingless optical fiber. Both silica and plastic optical fibers are used as a substrate. Such a fiber converts light modes propagating in the modified cladding into electrical signal. DSC structure consisting of ZnO:Al transparent current collector layer, TiO₂ photoelectrode sensitized with ruthenium dye, gelatinized iodine electrolyte, and carbon-based counter electrode was deposited layer by layer on top of the optical fiber. Current density-voltage curves of photovoltaic (PV) fibers of different diameters are presented. Maximum obtained short circuit current, I_sc, was 26 nA/cm² and maximum open circuit voltage, V_oc, was 0.44 V. The fabrication issues and applications of the PV fiber are discussed in the article.     

Doping-free fabrication of carbon nanotube thin-film diodes and their photovoltaic characteristics

Random networks of single-walled carbon nanotubes (SWCNTs) were have been grown by chemical vapor deposition on silicon wafers and used for fabricating field-effect transistors (FETs) using symmetric Pd contacts and diodes using asymmetrical Pd and Sc contacts. For a short channel FET or diode with a channel length of about 1 μm or less, the device works in the direct transport regime, while for a longer channel device the transport mechanism changes to percolation. Detailed electronic and photovoltaic (PV) characterizations of these carbon nanotube (CNT) thin-film devices was carried out. While as-fabricated FETs exhibited typical p-type transfer characteristics, with a large current ON/OFF ratio of more than 10⁴ when metallic CNTs were removed via a controlled breakdown, it was found that the threshold voltage for the devices was typically very large, of the order of about 10 V. This situation was greatly improved when the device was coated with a passivation layer of 12 nm HfO₂, which effectively moved the threshold voltages of both FET and diode back to center around zero or turned these device to their OFF states when no bias was applied on the gate. PV measurements were then made on the short channel diodes under infrared laser illumination. It was shown that under an illumination power density of 1.5 kW/cm², the device resulted in an open circuit voltage V _OC = 0.21 V and a short circuit current I _SC = 3.74 nA. Furthermore, we compared PV characteristics of CNT film diodes with different channel lengths, and found that the power transform efficiency decreased significantly when the device changed from the direct transport to the percolation regime.     

14 kV single-phase superconducting fault current limiter based on YBCO films

We present the fabrication and short circuit test results of a 14 kV single-phase resistive superconducting fault current limiter (SFCL) based on YBa₂Cu₃O₇ (YBCO) films. Individual components were processed using the 4″ YBCO films and have the rated voltage and current of 600 V and 35 A at 77 K, respectively. Twenty four components, eight components in series and three lines in parallel, make a module having the rated voltage and current of 4.8 kV and 105 A, respectively. Three modules were assembled in series to produce the SFCL working at 77 K, a 14 kV single-phase machine for the 22.9 kV Y-Y grid. short circuit tests were successfully conducted in an accredited test facility with the maximum fault currents up to 14.1 kA_P. All components quenched together upon faults and shared the rated voltage evenly without any supplementary device between the modules. This proves that the SFCL based on YBCO films may not only work reliably at 22.9 kV, but also provide technical feasibility for higher voltage application including the transmission grids.     

基于EMD模糊熵和SVM的转子系统故障诊断

提出一种经验模态分解、模糊熵和支持向量机相结合的转子系统故障诊断方法。该方法首先对转子系统故障信号进行经验模态分解,得到若干阶表征故障信息的固有模态函数,并运用基于能量原理的虚假模态消除方法剔除虚假模态分量;再利用模糊熵能够表示信号复杂程度且具有相对稳定性等特点,选取前4阶固有模态函数的模糊熵值作为各故障信号的特征向量;最后将该特征向量输入到支持向量机中进行转子系统的故障分类。试验结果表明,该方法能够有效的提取转子系统故障特征和对转子系统进行故障诊断。