基于MHMM-SVM混合模型的雷达发射机性能退化状态监测

雷达发射机结构复杂,早期性能退化状态呈现多样化。为此,给出了一种MHMM-SVM故障诊断模型。该模型首先通过小波分析来提取观测信号的特征向量,然后从提取的特征向量中提取部分作为样本完成模型训练并生成发射机各种退化状态的分类器,最后给出诊断结果。仿真验证结果表明:MHMM-SVM故障诊断模型对电子装备早期故障的识别率明显优于MHMM、SVM单一模型。     

关于半鞅的可料表示性

利用鞅空间H1的泛函表示定理、泛函分析中的Hahn-Banach定理、半鞅向量随机积分的Girsanov定理,获得了半鞅可料表示性的特征。由于使用的是半鞅的向量随机积分,它推广了经典的结论。     

从图像中提取文字

提出一种从图像中检测目标的思路 ,并利用它提出了从图像中提取文字的新方法。该方法可以用于图像数据库内容检索 ,视频检索 ,以及将纸张出版物中的文本转化为电子出版物。新方法分为如下两个基本步骤 ,首先检测出可能存在文字的候选区 ;其次 ,利用Hough变换方法分析候选内扫描线灰度分布对候选区进行证实 ,并对候选区的形状进行修剪 ,使其尽可能紧凑地包含文字。该方法适用于检测出不同大小、字体、排列方向、灰度 (颜色 )的文字。由于该方法所有操作都是基于像元进行的 ,所以便于硬件并行实现。所做的实验证实了方法的有效性     

一种利用小波包函数调制的 MC-CDMA 及其性能

将小波包变换理论应用于多载波码分多址(MC-CDMA)系统,提出一种基于小波包函数的正交调制方法.系统中,发端数据串并转换成若干子信道,每个子信道的数据进行直序扩频,不同子信道上的数据以chip调制不同的小波包函数作为基带传输波形.小波包函数的一系列优良性质,使系统能有效地抗多径干扰,抑制子信道串扰、多址干扰等.本文对系统在多径Rayleigh慢衰落信道中的性能,进行了理论分析和计算机仿真.结果表明,本系统比基于DFT的MC-CDMA有较大的性能改善.     

A novel signal feature extraction technology based on empirical wavelet transform and reverse dispersion entropy

Feature extraction is an important part of signal processing, which is significant for signal detection, classification, and recognition. The nonlinear dynamic analysis method can extract the nonlinear characteristics of signals and is widely used in different fields. Reverse dispersion entropy (RDE) proposed by us recently, as a nonlinear dynamic analysis method, has the advantages of fast computing speed and strong anti-noise ability, which is more suitable for measuring the complexity of signal than traditional permutation entropy (PE) and dispersion entropy (DE). Empirical wavelet transform (EWT), based on the theory of wavelet analysis, can decompose a complex non-stationary signal into a number of empirical wavelet functions (EWFs) with compact support set spectrum, which has better decomposition performance than empirical mode decomposition (EMD) and its improved algorithms. Considering the advantages of RDE and EWT, on the one hand, we introduce EWT into the field of underwater acoustic signal processing and fault diagnosis to improve the signal decomposition accuracy; on the other hand, we use RDE as the features of EWFs to improve the signal separability and stability. Finally, we propose a novel signal feature extraction technology based on EWT and RDE in this paper. Experimental results show that the proposed feature extraction technology can effectively extract the complexity features of actual signals. Moreover, it also has higher distinguishing ability for different types of signals than five latest feature extraction technologies.     

改进的四陀螺冗余捷联惯组故障诊断与隔离方法

四个陀螺和四个加速度计构成的冗余捷联惯组系统,采用等价空间法——广义似然比法、最优奇偶向量法进行故障诊断与隔离时,会出现不能正确隔离出故障仪表的问题。分析后得出,存在一个余度时,等价空间法只能诊断出故障,而无法隔离故障仪表,这是由等价空间法的原理决定的。小波变换可对信号进行多尺度分析,能够准确地找出信号奇异点——幅值突变和频率突变等。基于以上问题,利用小波变换对突变信号敏感的特性,提出等价空间+小波变换的组合方法,对四陀螺冗余捷联惯组进行故障诊断与隔离。数字仿真试验证明:对于阶跃型故障,该组合方法能诊断故障并能成功进行故障隔离。     

Minimal LST representations of MAP(n)s: Moment fittings and queueing approximations

A Markovian arrival process of order n, MAP(n), is typically described by two n × n transition rate matrices in terms of rate parameters. While it is straightforward and intuitive, the Markovian representation is redundant since the minimal number of parameters is n² for non‐redundant MAP(n). It is well known that the redundancy complicates exact moment fittings. In this article, we present a minimal and unique Laplace‐Stieltjes transform (LST) representations for MAP(n)s. Even though the LST coefficients vector itself is not a minimal representation, we show that the joint LST of stationary intervals can be represented with the minimum number of parameters. We also propose another minimal representation for MAP(3)s based on coefficients of the characteristic polynomial equations of the two transition rate matrices. An exact moment fitting procedure is presented for MAP(3)s based on two proposed minimal representations. We also discuss how MAP(3)/G/1 departure process can be approximated as a MAP(3). A simple tandem queueing network example is presented to show that the MAP(3) performs better than the MAP(2) in queueing approximations especially under moderate traffic intensities. © 2016 Wiley Periodicals, Inc. Naval Research Logistics 63: 549–561, 2016     

一种基于改进SIFT的图像盲取证算法

针对常见的复制粘贴篡改,提出了一种基于改进SIFT（Scale Invariant Feature Transform）的图像篡改检测算法。首先,把待检测图像划分成多个矩形的ROS（Regions of Suspicion）区域,利用改进的SIFT算法提取各个ROS区域的关键点;然后,利用马氏距离原理对特征点对进行筛选与匹配,通过直线将这些匹配点连接起来,对复制粘贴区域进行准确定位。实验表明,该算法不仅降低了运算复杂度,而且对篡改区域的平移、缩放,旋转、亮度调整等,都有很好的检测和定位效果。     

基于shearlet与PCNN的多聚焦图像融合方法

针对传统脉冲耦合神经网络(PCNN)模型在多聚焦图像融合领域应用中面临的参数繁杂等问题,提出一种基于剪切波(shearlet)变换与改进型PCNN的多聚焦图像融合方法。相比以往的变换域方法,shearlet具有理想的图像信息捕捉性能以及较低的计算复杂度,因此,可利用shearlet将待融合图像进行多尺度多方向分解。其次,对经典PCNN模型加以改进,综合运用清晰度水平以及协调矩阵完成低频子带图像以及一系列高频子带图像的融合过程。最后,运行shearlet反变换得到最终融合图像。仿真实验选取了若干组待融合图像进行仿真,验证了该方法在主、客观评价两方面的优越性。     

基于不同能量值特征优选的水声目标识别

采用Bartlet平均周期图、小波变换和经验模态分解方法分别提取信号的频段能量值特征、IMF能量值特征,重点对IMF能量值特征进行特征优选。通过设计BP神经网络分类器,对实测的四类舰船目标的辐射噪声信号进行测试,取得了较好的识别效果。