Decentralized M-ary detection via hierarchical binary decision fusion

We consider the distributed M-ary detection problem. The M-ary decision-making process is implemented via a sequence of binary decision-making processes. The resulting binary decisions represent a hierarchical partition of the M-ary object space, which is organized in the form of a binary decision tree. This approach breaks a complex M-ary decision fusion problem into a set of much simpler binary decision fusion problems. We first develop a method for partitioning the M-ary object space. We then obtain the optimal decision rules that the fusion center and the sensors employ at the internal nodes of the binary decision tree. The results are illustrated in an example.     

Multi-focus image fusion based on sparse decomposition and background detection

The goal of image fusion is to accurately and comprehensively describe complementary information of multiple source images in a new scene. Traditional fusion methods are easy to produce side-effects which cause artifacts and blurred edges. To solve these problems, a novel fusion algorithm based on robust principal component analysis (RPCA) and guided filter is proposed. The guided filter can preserve the edges effectively, which is often used to enhance the images without distort the details. Considering edges and flat area are treated differently by the guided filter, in this paper, sparse component of the source image is filtered by the guided filter to generate the enhanced image which contains the preserved edges and the enhanced background. And then the focused regions of the source images are detected by spatial frequency map of the difference images between the enhanced image and the corresponding source image. Finally, morphological algorithm is used to obtain precise fusion decision map. Experimental results show that the proposed method improves the fusion performance obviously which outperforms the current fusion methods.     

无人机PCA故障检测与诊断技术研究

无人机（UAV）飞控系统传感器故障检测和诊断常采用基于解析模型的方法,但飞行控制系统（FCS）的精确数学模型往往获取困难。针对此问题,提出了一种UAV-PCA算法。该算法在传统主成分分析（PCA）方法的基础上结合方差敏感自适应阈值的故障检测方法和基于特征方向的故障诊断方法,实现UAV飞控系统传感器的故障检测和诊断。算法不需要系统的数学模型,解决了应用传统PCA方法进行FCS故障检测与诊断时易出现暂态过程虚警和误诊断的问题。仿真结果证明该算法可以快速准确地检测传感器故障,而且可以有效地降低暂态过程虚警和提高诊断结果准确度。     

基于图像分解和稀疏表示的多聚焦图像融合

针对多聚焦图像融合中缺乏细节保护和结构不连续的不足,提出了一种基于图像分解的多聚焦图像融合算法.首先,源图像采用卡通纹理图像分解得到卡通部分和纹理部分;其次,卡通部分采用卷积稀疏表示的方法进行融合,纹理部分采用字典学习进行融合;最后,将卡通和纹理部分融合得到融合图像.实验建立在标准的融合数据集中,并与传统和最近的融合方法进行比较.实验结果证明,该算法所获得的融合结果在方差和信息熵上具有更好的表现,该算法能够有效克服多聚焦图像融合中缺乏细节保护和结构不连续的缺点,同时有更好的视觉效果.     

基于PCA和ELM的模拟电路故障诊断

针对模拟电路的故障诊断和健康管理（PHM）的应用,提出了结合主成分分析（PCA）和极限学习机（ELM）的故障诊断方法。该方法用Sallen-Key带通滤波器来获取故障样本,并通过PCA进行故障特征提取。根据故障样本对ELM进行训练来获得故障诊断模型。实验结果表明,该实现方法识别率高、鲁棒性好,在工程实际中具有研究和应用价值。     

Kernel PCA与BP神经网络相结合的变压器故障诊断*

为了提高变压器故障诊断的准确率和抗干扰能力,提出一种基于核特征量的BP神经网络故障诊断模型。通过核主成分分析将故障样本从低维的特征空间非线性地映射到高维的核空间,提高了样本的可分性,然后以核特征量作为BP神经网络的输入特征量,建立变压器故障诊断模型。实验对比了结构相似、输入量不同的BP神经网络,结果表明采用核特征量的诊断模型具有更好的诊断效果和抗干扰能力。     

基于边缘保留分解和改进稀疏表示的医学图像融合

针对多模态医学图像融合中容易产生伪影且存在细节缺失的问题,提出一种利用多尺度边缘保留分解和稀疏表示的二尺度多模态医学图像融合方法框架.首先利用边缘保留滤波器对源图像进行多尺度分解,得到源图像的平滑层和细节层.然后,将改进的稀疏表示算法用于融合平滑层,并在此基础上提出一种基于图像块筛选的策略来构建过完备字典的数据集,再利...     

基于主元分析的桥梁挠度传感器故障诊断研究

主元分析(PCA)是一种典型的数据降维的多元统计方法,已被越来越多地用于故障诊断。将PCA应用在桥梁挠度传感器故障诊断。介绍了PCA的理论,研究了基于PCA的故障检测方法和基于贡献率的故障诊断方法。计算平方预测误差(SPE)和Hoteling T2统计,当统计量超过阈值时,判断系统出现了传感器故障,然后通过SPE贡献图判断故障源。通过仿真验证了PCA在故障诊断的实用性,但结果也表明:PCA对小故障不是很敏感。     

基于低秩稀疏分解的自适应运动目标检测算法

传统的低秩稀疏分解方法使用[l1]范数把场景中的运动目标建模为稀疏离群值,分离出低秩的背景成分与稀疏的运动目标成分。然而,在许多实际场景中往往会有动态背景的情形（例如水面波纹、树木摇动）,[l1]范数并不能区分出这些干扰与真实目标,从而大大影响检测效果。实际上,运动目标区域中的像素不仅仅具有稀疏性,还具有空间分布上的连续性。通过引入空间融合稀疏约束,在空间连续性和稀疏性两方面对运动目标进行建模,使模型更符合目标像素的分布规律。同时,设计了一种自适应的参数更新方法,使算法的鲁棒性进一步提升。在公共数据集上的大量实验表明,相比于传统方法,该算法在准确率和鲁棒性方法有很大提高。     

Intelligent fault diagnosis of turbine blade cracks via multiscale sparse filtering and multi-kernel support vector machine for information fusion

For accurately identifying the crack severity of turbine blades, a novel intelligent diagnosis framework is proposed in our paper, which uses multiscale sparse filtering (MSF)-based unsupervised sparse feature learning and multi-kernel support vector machine for information fusion (MKSVMIF). To realize the technical readiness and the state-of-the-art diagnostic performance, advanced signal processing methods are used to eliminate interference and retain fault-related characteristics. Thus, the EEMD-based multiwavelet packet energy entropy (EEMD-WPEE) as an enhanced method is proposed for multi-source three-dimensional blade tip clearance signals (3D-BTC), which is used to enhance the fault-related information. Afterward, the MSF is constructed to adaptively learn multisource sparse features from EEMD-WPEE representation of 3D-BTC by an unsupervised manner. Finally, the MKSVMIF is proposed to fuse these sparse features and diagnose crack severities. To validate the effectiveness of our proposed MSF-MKSVMIF framework, Extensive experiments are conducted on a blade-rotor simulation rig, and the results show that our proposed framework is superior to other comparison methods and can quantitatively detect different blade crack severities with a relatively small number of samples.     

基于故障检测次序的供料单元故障诊断

针对气动PLC自动生产线中供料单元,在一次供料过程时,上电后却无法运作,通过观察其故障现象,分析其故障原因,提出设定故障检查次序,综合利用假设验证法、替换法、经验法和测量法等故障诊断方法,排除设备的故障,继而通过实践证明合理设定故障检查次序对设备故障排除的重要性.     

Decision-level fusion based on wavelet decomposition for induction motor fault diagnosis using transient current signal

In this paper, we propose and implement a decision-level fusion model by combining the information of multi-level wavelet decomposition for fault diagnosis of induction motor using transient stator current signal. Firstly, the start-up transient current signals are collected from different faulty motors. Then signal preprocessing is conducted containing smoothing and subtracting to reduce the influence of line frequency in transient current signals. Next, we employ discrete wavelet transform technique to decompose the preprocessed signals into different frequency ranges of products, and then features are extracted from decomposed detail components. Finally, two decision-level fusion strategies, Bayesian belief fusion and multi-agent fusion, are employed. That is, fault features are classified using several classifiers and generated decisions are fused using a specific fusion algorithm. The proposed approach is evaluated by an experiment of fault diagnosis for induction motors. Experiment results show that excellent diagnosis performance can be obtained.     

自适应融合局部和全局稀疏表示的图像显著性检测

针对基于局部或全局稀疏表示的图像显著性检测方法频繁出现提取对象不完整、边界不光滑及噪声消除不干净等问题,提出自适应融合局部和全局稀疏表示的图像显著性检测方法。首先,对原始图像进行分块处理,利用图像块代替像素操作,降低算法复杂度;其次,对分块后的图像进行局部稀疏表示,即：针对每一个图像块,选取其周围的若干图像块生成过完备字典,基于该字典对图像块进行稀疏重构,得到原始图像的初始局部显著图,该显著图能够有效提取显著性目标的边缘信息;接着,对分块后的图像进行全局稀疏表示,与局部稀疏表示过程类似,不同的是针对每一个图像块所生成的字典来源于图像四周边界处的图像块,这样可以得到能有效检测出显著性目标内部区域的初始全局显著图;最后,将初始局部和全局显著图进行自适应融合,生成最终显著图。实验结果表明,提出算法在查准率（precision）、查全率（recall）及F-measure等指标上优于几种经典的图像显著性检测方法。     

基于信息融合技术的电机故障诊断

为了能够从多方面反映电机系统状态,实现对电机故障模式的自动识别与准确诊断,将信息融合技术与神经网络相结合,建立电机故障诊断系统。在数据融合级上,将故障特征量进行分类处理,然后,采用多层神经网络进行故障特征级融合与电机故障的局部诊断,获得彼此独立的证据,再运用DempserShafer(D-S)证据理论融合算法对各证据进行融合,最终,实现对电机故障的准确诊断。诊断测试试验证明:该诊断系统提高了电机故障诊断的精度,并能满足诊断的实时性要求。     

基于概率密度PCA的多模态过程故障检测

为了提高故障检测和分类能力,提出基于概率密度PCA的多模态过程故障检测算法。对各模态的训练数据建立PCA模型,计算各个模型的控制限和匹配系数。根据匹配系数计算各模态统一的控制限。对新来的数据,运用概率密度确定其模态。新来数据向对应模态的模型上投影并计算统一的统计量,比较统计量与控制限进行多模态过程故障检测。把该方法应用到数值例子和半导体过程中,仿真结果表明,该算法在分类及多模态过程故障检测方面具有很高的准确性。     

基于特征子空间的滑动窗PCA在批过程故障诊断中的应用

基于传统的多向主元分析MPCA(multiway principal component analysis)常会导致误诊断,且对批过程难以保证在线状态监测和故障诊断的实时性,提出了一种基于特征子空间的滑动窗主元分析方法。在实时故障监测与诊断时,该方法采用适当大小的滑动窗逐步更新当前子数据空间,对当前子数据空间故障的识别通过依次计算其与基底库中各故障的匹配度来进行。这种方法克服了传统的MPCA不能处理非线性过程和实时性问题,并避免了MPCA在线应用时预报未来测量值带来的误差, 提高了批过程性能监测和故障诊断的准确性。     

基于决策树和相关向量机的智能故障诊断方法

针对故障诊断面临的故障样本少、非线性强、多故障处理等问题以及传统智能诊断方法存在的不足,提出了一种基于决策树（DT）和相关向量机（RVM）的智能故障诊断方法。通过构造决策二叉树,将多类分类问题分解成多个二类分类问题;在各个决策节点,利用RVM进行二类分类,从而实现RVM的多类分类。理论分析及仿真结果表明,相比支持向量机,新方法在保持高诊断正确率的同时具有更高的稀疏性和诊断效率,并且能够提供概率式输出,更具实用价值;相比OAR-RVM和OAO-RVM方法,新方法节省了训练时间,具有更高的训练效率。     

一种新的神经网络故障诊断方法

基于数据融合思想，提出一种新的神经网络故障诊断方法。利用系统故障征兆的分散性和复杂性，采用多个神经网络分别对每一类故障进行诊断，网络输入为与输出故障相关联的监测信号的特征值，将各网络输出进行融合，给出最后诊断结果。将该方法应用于斜轴式无铰柱塞液压泵故障诊断，结果表明能够充分利用各种特征信息，提高诊断速度和精确度。