In this paper, we introduce the definition of L-fuzzy vector subspace, define its dimension by an L-fuzzy natural number. For a finite-dimensional L-fuzzy vector subspace, we prove that the equality holds without any ...In this paper, we introduce the definition of L-fuzzy vector subspace, define its dimension by an L-fuzzy natural number. For a finite-dimensional L-fuzzy vector subspace, we prove that the equality holds without any restricted conditions. At the same time, we deduce that the formula holds.展开更多
The vector sampling theorem has been investigated and widely used by multi-channel deconvolution, multi-source separation and multi-input multi-output （MIh40） systems. Commonly, for most of the results on MIMO syste...The vector sampling theorem has been investigated and widely used by multi-channel deconvolution, multi-source separation and multi-input multi-output （MIh40） systems. Commonly, for most of the results on MIMO systems, the input signals are supposed to be band-limited. In this paper, we study the vector sampling theorem for the wavelet subspaces with reproducing kernel. The case of uniform sampling is discussed, and the necessary and sufficient conditions for reconstruction are given. Examples axe also presented.展开更多
Let F be any commutative field. Let v be an integer≥1 and be a fixed 2v × 2v nonsingular alternate matrix over F. Define Sp(F)={T: 2v×2v matrix over F|TKT~T=K}. It is well-known that Sp(F) is a group with r...Let F be any commutative field. Let v be an integer≥1 and be a fixed 2v × 2v nonsingular alternate matrix over F. Define Sp(F)={T: 2v×2v matrix over F|TKT~T=K}. It is well-known that Sp(F) is a group with respect to the matrix multiplication and is called the symplectic group of degree 2v over F展开更多
Hyperspectral images in remote sensing include hundreds of spectral bands that provide valuable information for accurately identify objects.In this paper,a new method of classifying hyperspectral images using spectral...Hyperspectral images in remote sensing include hundreds of spectral bands that provide valuable information for accurately identify objects.In this paper,a new method of classifying hyperspectral images using spectral spatial information has been presented.Here,using the hyperspectral signal subspace identification(HYSIME)method which estimates the signal and noise correlation matrix and selects a subset of eigenvalues for the best representation of the signal subspace in order to minimize the mean square error,subsets from the main sample space have been extracted.After subspace extraction with the help of the HYSIME method,the edge-preserving filtering(EPF),and classification of the hyperspectral subspace using a support vector machine(SVM),results were then merged into the decision-making level using majority rule to create the spectral-spatial classifier.The simulation results showed that the spectral-spatial classifier presented leads to significant improvement in the accuracy and validity of the classification of Indiana,Pavia and Salinas hyperspectral images,such that it can classify these images with 98.79%,98.88% and 97.31% accuracy,respectively.展开更多
Millimeter-wave communication （mmWC） is considered as one of the pioneer candidates for 5G indoor and outdoor systems in E-band. To subdue the channel propagation characteristics in this band, high dimensional anten...Millimeter-wave communication （mmWC） is considered as one of the pioneer candidates for 5G indoor and outdoor systems in E-band. To subdue the channel propagation characteristics in this band, high dimensional antenna arrays need to be deployed at both the base station （BS） and mobile sets （MS）. Unlike the conventional MIMO systems, Millimeter-wave （mmW） systems lay away to employ the power predatory equipment such as ADC or RF chain in each branch of MIMO system because of hardware constraints. Such systems leverage to the hybrid precoding （combining） architecture for downlink deployment. Because there is a large array at the transceiver, it is impossible to estimate the channel by conventional methods. This paper develops a new algorithm to estimate the mmW channel by exploiting the sparse nature of the channel. The main contribution is the representation of a sparse channel model and the exploitation of a modified approach based on Multiple Measurement Vector （MMV） greedy sparse framework and subspace method of Multiple Signal Classification （MUSIC） which work together to recover the indices of non-zero elements of an unknown channel matrix when the rank of the channel matrix is defected. In practical rank-defective channels, MUSIC fails, and we need to propose new extended MUSIC approaches based on subspace enhancement to compensate the limitation of MUSIC. Simulation results indicate that our proposed extended MUSIC algorithms will have proper performances and moderate computational speeds, and that they are even able to work in channels with an unknown sparsity level.展开更多
In the second paper on the inverse relativity model, we explained in the first paper [1] that analyzing the four-dimensional displacement vector on space-time according to a certain approach leads to the splitting of ...In the second paper on the inverse relativity model, we explained in the first paper [1] that analyzing the four-dimensional displacement vector on space-time according to a certain approach leads to the splitting of space-time into positive and negative subspace-time. Here, in the second paper, we continue to analyze each of the four-dimensional vectors of velocity, acceleration, momentum, and forces on the total space-time fabric. According to the approach followed in the first paper. As a result, in the special case, we obtain new transformations for each of the velocity, acceleration, momentum, energy, and forces specific to each subspace-time, which are subject to the positive and negative modified Lorentz transformations described in the first paper. According to these transformations, momentum remains a conserved quantity in the positive subspace and increases in the negative subspace, while the relativistic total energy decreases in the positive subspace and increases in the negative subspace. In the general case, we also have new types of energy-momentum tensor, one for positive subspace-time and the other for negative subspace-time, where the energy density decreases in positive subspace-time and increases in negative subspace-time, and we also obtain new gravitational field equations for each subspace-time.展开更多
在说话人识别研究中,基于身份认证矢量(identity vector,i-vector)的子空间建模被证明是目前最前沿最有效的说话人建模技术,其中如何有效准确地估计总体变化子空间矩阵T成为影响系统性能好坏的关键问题.本文针对i-vector技术如何在新的...在说话人识别研究中,基于身份认证矢量(identity vector,i-vector)的子空间建模被证明是目前最前沿最有效的说话人建模技术,其中如何有效准确地估计总体变化子空间矩阵T成为影响系统性能好坏的关键问题.本文针对i-vector技术如何在新的应用环境下进行总体变化子空间矩阵T的自适应估计问题进行了研究,并提出了两种行之有效的自适应估计算法.在由美国国家标准技术局(American National Institute of Standard and Technology,NIST)组织的2008年说话人识别核心评测数据库以及自行采集的测试数据库上的实验结果显示,不论采用测试集数据本身还是与测试集较匹配的开发集数据,通过本文所提的自适应算法来更新总体变化子空间矩阵均可以使更新后的子空间更有利于新测试数据下的低维子空间描述,在新的测试环境下都更有利于说话人分类.此外实验结果还表明基于多子空间拼接的子空间自适应方法性能明显优于迭代自适应方法,而且两者的结合可达到最优的识别性能,且此时利用开发集数据进行自适应可以接近其利用测试集数据进行自适应得到的最优性能.展开更多
An important problem that arises in different areas of science and engineering is that of computing the limits of sequences of vectors , where , N being very large. Such sequences arise, for example, in the solution o...An important problem that arises in different areas of science and engineering is that of computing the limits of sequences of vectors , where , N being very large. Such sequences arise, for example, in the solution of systems of linear or nonlinear equations by fixed-point iterative methods, and are simply the required solutions. In most cases of interest, however, these sequences converge to their limits extremely slowly. One practical way to make the sequences converge more quickly is to apply to them vector extrapolation methods. Two types of methods exist in the literature: polynomial type methods and epsilon algorithms. In most applications, the polynomial type methods have proved to be superior convergence accelerators. Three polynomial type methods are known, and these are the minimal polynomial extrapolation (MPE), the reduced rank extrapolation (RRE), and the modified minimal polynomial extrapolation (MMPE). In this work, we develop yet another polynomial type method, which is based on the singular value decomposition, as well as the ideas that lead to MPE. We denote this new method by SVD-MPE. We also design a numerically stable algorithm for its implementation, whose computational cost and storage requirements are minimal. Finally, we illustrate the use of SVD-MPE with numerical examples.展开更多
针对传统多重信号分类(multiple signal classification,MUSIC)算法在低信噪比环境和小型化麦克风阵列影响下的性能下降问题,提出了一种结合第一主向量法和子空间加权法的改进MUSIC算法。首先利用第一主向量法对传统MUSIC算法进行优化,...针对传统多重信号分类(multiple signal classification,MUSIC)算法在低信噪比环境和小型化麦克风阵列影响下的性能下降问题,提出了一种结合第一主向量法和子空间加权法的改进MUSIC算法。首先利用第一主向量法对传统MUSIC算法进行优化,得到改进的空间谱函数,以降低噪声对定位精度的影响:其次利用基于双指数模型的最小二乘法修正特征值,并对信号子空间和噪声子空间进行加权处理。仿真结果表明,改进后的MUSIC算法能够有效提升小型化麦克风阵列在低信噪比条件下对相近声源波达方向的估计精度,为声源定位系统的小型化应用提供了新的解决方案。展开更多
In order to accurately identify a bearing fault on a wind turbine, a novel fault diagnosis method based on stochastic subspace identification(SSI) and multi-kernel support vector machine(MSVM) is proposed. Firstly, th...In order to accurately identify a bearing fault on a wind turbine, a novel fault diagnosis method based on stochastic subspace identification(SSI) and multi-kernel support vector machine(MSVM) is proposed. Firstly, the collected vibration signal of the wind turbine bearing is processed by the SSI method to extract fault feature vectors. Then, the MSVM is constructed based on Gauss kernel support vector machine(SVM) and polynomial kernel SVM. Finally, fault feature vectors which indicate the condition of the wind turbine bearing are inputted to the MSVM for fault pattern recognition. The results indicate that the SSI-MSVM method is effective in fault diagnosis for a wind turbine bearing and can successfully identify fault types of bearing and achieve higher diagnostic accuracy than that of K-means clustering, fuzzy means clustering and traditional SVM.展开更多
In order to ease the pass-band response distortion of the matrix pre-filter,a simple approach for designing matrix spatial filter is proposed,which minimizes the sum of the k maximal distortion norm（k is the number o...In order to ease the pass-band response distortion of the matrix pre-filter,a simple approach for designing matrix spatial filter is proposed,which minimizes the sum of the k maximal distortion norm（k is the number of the constraint points）within the pass-band,while constraining the filter response within the stop-band.Considering the costly amount of calculation of the high-resolution methods,an algorithm with small amount of calculation based on matrix pre-filtering and subspace fitting using acoustic vector array（MF-VSSF）is proposed.Through joint processing of signal subspace of both pressure and particle velocity,the pre-filtering matrix and the signal subspace is decreased to M-dimensional（M is the number of array-element）,hence reduces the time-consumption of the matrix pre-filter design and DOA searching.Simulation results show that,the method offers the same performance as MUSIC with pre-filtering,but has much lesser amount of calculation.Moreover,the designed prefilter can efficiently suppress the interference in the stop-band and improve the estimation and resolution performance of successive DOA estimators.展开更多
特征迁移重在领域共有特征间学习,然而其忽略领域特有特征的判别信息,使算法的适应性受到一定的局限.针对此问题,提出了一种融合异构特征的子空间迁移学习(The subspace transfer learning algorithm integrating with heterogeneous fe...特征迁移重在领域共有特征间学习,然而其忽略领域特有特征的判别信息,使算法的适应性受到一定的局限.针对此问题,提出了一种融合异构特征的子空间迁移学习(The subspace transfer learning algorithm integrating with heterogeneous features,STL-IHF)算法.该算法将数据的特征空间看成共享和特有两个特征子空间的组合,同时基于经验风险最小框架将共享特征和特有特征共同嵌入到支持向量机(Support vector machine,SVM)的训练过程中.其在共享特征子空间上实现知识迁移的同时兼顾了领域特有的异构信息,增强了算法的适应性.模拟和真实数据集上的实验结果表明了所提方法的有效性.展开更多
文摘In this paper, we introduce the definition of L-fuzzy vector subspace, define its dimension by an L-fuzzy natural number. For a finite-dimensional L-fuzzy vector subspace, we prove that the equality holds without any restricted conditions. At the same time, we deduce that the formula holds.
基金supported by the National Natural Science Foundation of China (Grant No.60873130)the Shanghai Leading Academic Discipline Project (Grant No.J50104)
文摘The vector sampling theorem has been investigated and widely used by multi-channel deconvolution, multi-source separation and multi-input multi-output （MIh40） systems. Commonly, for most of the results on MIMO systems, the input signals are supposed to be band-limited. In this paper, we study the vector sampling theorem for the wavelet subspaces with reproducing kernel. The case of uniform sampling is discussed, and the necessary and sufficient conditions for reconstruction are given. Examples axe also presented.
文摘Let F be any commutative field. Let v be an integer≥1 and be a fixed 2v × 2v nonsingular alternate matrix over F. Define Sp(F)={T: 2v×2v matrix over F|TKT~T=K}. It is well-known that Sp(F) is a group with respect to the matrix multiplication and is called the symplectic group of degree 2v over F
文摘Hyperspectral images in remote sensing include hundreds of spectral bands that provide valuable information for accurately identify objects.In this paper,a new method of classifying hyperspectral images using spectral spatial information has been presented.Here,using the hyperspectral signal subspace identification(HYSIME)method which estimates the signal and noise correlation matrix and selects a subset of eigenvalues for the best representation of the signal subspace in order to minimize the mean square error,subsets from the main sample space have been extracted.After subspace extraction with the help of the HYSIME method,the edge-preserving filtering(EPF),and classification of the hyperspectral subspace using a support vector machine(SVM),results were then merged into the decision-making level using majority rule to create the spectral-spatial classifier.The simulation results showed that the spectral-spatial classifier presented leads to significant improvement in the accuracy and validity of the classification of Indiana,Pavia and Salinas hyperspectral images,such that it can classify these images with 98.79%,98.88% and 97.31% accuracy,respectively.
文摘Millimeter-wave communication （mmWC） is considered as one of the pioneer candidates for 5G indoor and outdoor systems in E-band. To subdue the channel propagation characteristics in this band, high dimensional antenna arrays need to be deployed at both the base station （BS） and mobile sets （MS）. Unlike the conventional MIMO systems, Millimeter-wave （mmW） systems lay away to employ the power predatory equipment such as ADC or RF chain in each branch of MIMO system because of hardware constraints. Such systems leverage to the hybrid precoding （combining） architecture for downlink deployment. Because there is a large array at the transceiver, it is impossible to estimate the channel by conventional methods. This paper develops a new algorithm to estimate the mmW channel by exploiting the sparse nature of the channel. The main contribution is the representation of a sparse channel model and the exploitation of a modified approach based on Multiple Measurement Vector （MMV） greedy sparse framework and subspace method of Multiple Signal Classification （MUSIC） which work together to recover the indices of non-zero elements of an unknown channel matrix when the rank of the channel matrix is defected. In practical rank-defective channels, MUSIC fails, and we need to propose new extended MUSIC approaches based on subspace enhancement to compensate the limitation of MUSIC. Simulation results indicate that our proposed extended MUSIC algorithms will have proper performances and moderate computational speeds, and that they are even able to work in channels with an unknown sparsity level.
文摘In the second paper on the inverse relativity model, we explained in the first paper [1] that analyzing the four-dimensional displacement vector on space-time according to a certain approach leads to the splitting of space-time into positive and negative subspace-time. Here, in the second paper, we continue to analyze each of the four-dimensional vectors of velocity, acceleration, momentum, and forces on the total space-time fabric. According to the approach followed in the first paper. As a result, in the special case, we obtain new transformations for each of the velocity, acceleration, momentum, energy, and forces specific to each subspace-time, which are subject to the positive and negative modified Lorentz transformations described in the first paper. According to these transformations, momentum remains a conserved quantity in the positive subspace and increases in the negative subspace, while the relativistic total energy decreases in the positive subspace and increases in the negative subspace. In the general case, we also have new types of energy-momentum tensor, one for positive subspace-time and the other for negative subspace-time, where the energy density decreases in positive subspace-time and increases in negative subspace-time, and we also obtain new gravitational field equations for each subspace-time.
文摘在说话人识别研究中,基于身份认证矢量(identity vector,i-vector)的子空间建模被证明是目前最前沿最有效的说话人建模技术,其中如何有效准确地估计总体变化子空间矩阵T成为影响系统性能好坏的关键问题.本文针对i-vector技术如何在新的应用环境下进行总体变化子空间矩阵T的自适应估计问题进行了研究,并提出了两种行之有效的自适应估计算法.在由美国国家标准技术局(American National Institute of Standard and Technology,NIST)组织的2008年说话人识别核心评测数据库以及自行采集的测试数据库上的实验结果显示,不论采用测试集数据本身还是与测试集较匹配的开发集数据,通过本文所提的自适应算法来更新总体变化子空间矩阵均可以使更新后的子空间更有利于新测试数据下的低维子空间描述,在新的测试环境下都更有利于说话人分类.此外实验结果还表明基于多子空间拼接的子空间自适应方法性能明显优于迭代自适应方法,而且两者的结合可达到最优的识别性能,且此时利用开发集数据进行自适应可以接近其利用测试集数据进行自适应得到的最优性能.
文摘An important problem that arises in different areas of science and engineering is that of computing the limits of sequences of vectors , where , N being very large. Such sequences arise, for example, in the solution of systems of linear or nonlinear equations by fixed-point iterative methods, and are simply the required solutions. In most cases of interest, however, these sequences converge to their limits extremely slowly. One practical way to make the sequences converge more quickly is to apply to them vector extrapolation methods. Two types of methods exist in the literature: polynomial type methods and epsilon algorithms. In most applications, the polynomial type methods have proved to be superior convergence accelerators. Three polynomial type methods are known, and these are the minimal polynomial extrapolation (MPE), the reduced rank extrapolation (RRE), and the modified minimal polynomial extrapolation (MMPE). In this work, we develop yet another polynomial type method, which is based on the singular value decomposition, as well as the ideas that lead to MPE. We denote this new method by SVD-MPE. We also design a numerically stable algorithm for its implementation, whose computational cost and storage requirements are minimal. Finally, we illustrate the use of SVD-MPE with numerical examples.
文摘针对传统多重信号分类(multiple signal classification,MUSIC)算法在低信噪比环境和小型化麦克风阵列影响下的性能下降问题,提出了一种结合第一主向量法和子空间加权法的改进MUSIC算法。首先利用第一主向量法对传统MUSIC算法进行优化,得到改进的空间谱函数,以降低噪声对定位精度的影响:其次利用基于双指数模型的最小二乘法修正特征值,并对信号子空间和噪声子空间进行加权处理。仿真结果表明,改进后的MUSIC算法能够有效提升小型化麦克风阵列在低信噪比条件下对相近声源波达方向的估计精度,为声源定位系统的小型化应用提供了新的解决方案。
基金supported by National Key Technology Research and Development Program (No. 2015BAA06B03)
文摘In order to accurately identify a bearing fault on a wind turbine, a novel fault diagnosis method based on stochastic subspace identification(SSI) and multi-kernel support vector machine(MSVM) is proposed. Firstly, the collected vibration signal of the wind turbine bearing is processed by the SSI method to extract fault feature vectors. Then, the MSVM is constructed based on Gauss kernel support vector machine(SVM) and polynomial kernel SVM. Finally, fault feature vectors which indicate the condition of the wind turbine bearing are inputted to the MSVM for fault pattern recognition. The results indicate that the SSI-MSVM method is effective in fault diagnosis for a wind turbine bearing and can successfully identify fault types of bearing and achieve higher diagnostic accuracy than that of K-means clustering, fuzzy means clustering and traditional SVM.
基金supported by the National Natural Science Foundation of China(61201411)
文摘In order to ease the pass-band response distortion of the matrix pre-filter,a simple approach for designing matrix spatial filter is proposed,which minimizes the sum of the k maximal distortion norm（k is the number of the constraint points）within the pass-band,while constraining the filter response within the stop-band.Considering the costly amount of calculation of the high-resolution methods,an algorithm with small amount of calculation based on matrix pre-filtering and subspace fitting using acoustic vector array（MF-VSSF）is proposed.Through joint processing of signal subspace of both pressure and particle velocity,the pre-filtering matrix and the signal subspace is decreased to M-dimensional（M is the number of array-element）,hence reduces the time-consumption of the matrix pre-filter design and DOA searching.Simulation results show that,the method offers the same performance as MUSIC with pre-filtering,but has much lesser amount of calculation.Moreover,the designed prefilter can efficiently suppress the interference in the stop-band and improve the estimation and resolution performance of successive DOA estimators.
