Description: 不同算法对时间序列进行奇异谱分解的源程序代码，matlab2006a平台调试通过-different algorithms on time series singular spectral decomposition of the source code, matlab2006a through debugging platform Platform: |
Size: 6758 |
Author:mapleyu |
Hits:
Description: ************************************************************************ * * * * * THIS IS THE H Y P L A S 2.0 README FILE * * ----------------- * * * * HYPLAS is a finite element program for implicit small and large * * strain analisys of hyperelastic and elasto-plastic two-dimensional * * and axisymmetric solids * * * * HYPLAS v2.0 is the companion software to the textbook: * * EA de Souza Neto, D Peric & DRJ Owen. Computational Methods for * * Plasticity: Theory and Applications. Wiley, Chichester, 2008. * * (www.wiley.com/go/desouzaneto) * * * * Copyright (c) 1998-2008 EA de Souza Neto, D Peric, D.R.J. Owen * *----------------------------------------------------------------------* * File last updated: 18 October 2008 * * * * This file belongs in the directory ../HYPLAS_v2.0 * ************************************************************************ * * * I M P O R T A N T * * * * READ SECTIONS 0 TO 3 OF THIS FILE CAREFULLY BEFORE ATTEMPTING * * TO COMPILE AND RUN THE PROGRAM HYPLAS ON YOUR COMPUTER !! * * * * THE AUTHORS DO NOT GUARANTEE THAT ANY SUGGESTIONS/INSTRUCTIONS * * GIVEN IN THIS README FILE WILL WORK ON ANY PARTICULAR OPERATING * * SYSTEM. IF YOU DECIDE TO FOLLOW ANY SUGGESTIONS/INSTRUCTIONS * * GIVEN HERE YOU MUST DO SO AT YOUR OWN RISK. * * * * * * BUG REPORTS: Please send bug reports to * * * * hyplas_v2.0@live.co.uk * * * * Messages sent to the authors' personal email addresses * * will NOT be answered. * ************************************************************************ This file contains the following sections: 0. Copyright statement and disclaimer 0.(a) Copyright statement 0.(b) Disclaimer 0.(c) Conditions of use 1. Introduction 1.(a) Note on portability 2. Compiling and running HYPLAS 2.(a) Memory requirements 2.(b) Testing a newly compiled executable 3. The HYPLAS directory tree 4. Cross-referencing between the source code and the textbook 5. HYPLAS error messaging 6. Further remarks on HYPLAS ************************************************************************ 0. COPYRIGHT STATEMENT AND DISCLAIMER ================================== 0.(a) Copyright statement ------------------- You may only use this program for your own private purposes. You are not allowed, in any circumstances, to distribute this program (including its source code, executable and any other files related to it, either in their original version or any modifications introduced by you, the authors or any other party) in whole or in part, either freely or otherwise, in any medium, without the prior written consent of the copyright holders. 0.(b) Disclaimer ---------- This program (including its source code, executable and any other files related to it) is provided "as is" without warranty of any kind, either expressed or implied, including, but not limited to, any implied warranties of fitness for purpose. In particular, THIS PROGRAM IS BY NO MEANS GUARANTEED TO BE FREE FROM ERRORS. This program (or any modification incorporated to it by you, the authors or any other party) will run entirely at your risk. The results produced by this program are in no way guaranteed to be fit for any purpose. Under no circumstances will the authors/copyright holders be liable to anyone for damages, including any general, special, incidental or consequential damages arising from the use or inability to use the program (including, but not limited to, loss or corruption of data, failure of the program to operate in any particular way as well as damages arising from the use of any results produced by the program for any purpose). 0.(c) Conditions of use ----------------- You may only use this program if you fully understand and agree with the terms of the above disclaimer. You must not use this program if you do not agree with or do not understand (fully or in part) these conditions of use. 1. INTRODUCTION ============ HYPLAS is a finite element code for small and large strain analysis of hyperelastic and elasto-plastic solids. Most procedures implemented in HYPLAS are described in detail in its companion textbook: EA de Souza Neto, D Peric & DRJ Owen. Computational Methods for Plasticity: Theory and Applications. Wiley, Chichester, 2008 (www.wiley.com/go/desouzaneto). 1.(a) Note on Portability ------------------- HYPLAS has been written in standard ANSI FORTRAN 77. Currently, the only known (and deliberate) exceptions to the FORTRAN 77 ANSI standard are the instructions: INCLUDE '' used in many routines to include the HYPLAS database files (common blocks and global variables), and; CALL GETENV('HYPLASHOME',HYPLASHOME) used in subroutine "ERRPRT" (file ../HYPLAS_v2.0/src/GENERAL/errprt.f). This instruction inquires the name of the system environment variable HYPLASHOME and writes it on the character string HYPLASHOME. This instruction is NOT part of the ANSI FORTRAN 77 standard, but seems to work in most currently available FORTRAN 77 compilers. 2. COMPILING AND RUNNING H Y P L A S ================================== The HYPLAS source code is stored in directory ../HYPLAS_v2.0/src/ (../HYPLAS_v2.0/ being the current directory) and all its subdirectories. To generate an executable file, you just need to compile the FORTRAN source files: ../HYPLAS_v2.0/src/hyplas.f and ../HYPLAS_v2.0/src/*/*.f together. We recommend that the executable HYPLAS be stored in the directory ../HYPLAS_v2.0/bin to which the environment variable HYPLASHOME should be set (see below how to set a system environmental variable). WINDOWS (R) systems ------------------- On Microsoft Windows(R) systems, HYPLAS has been successfully compiled using Intel Visual Fortran Compiler(R) integrated with Microsoft Visual Studio(R). Here you only need to create a project that contains all Fortran source files mentioned above as well as the include files ..\HYPLAS_v2.0\src\*.INC On a Windows XP system, the system environment variable HYPLASHOME can be set as follows: 1. Open a File Manager 2. Right-click on the "My Computer" icon 3. Select "Properties" on the drop-down menu 4. A new window named "System Properties" will pop-up. Here select the "Advanced" tab. 5. On the "Advanced" tab, click the "Environment Variables" button. 6. A new window titled "Environment Variables" will pop-up. Here click the button "New" in the "System Variables" section of the window. 7. A new window will pop-up titled "New System Variable". Here you should fill the fields "Variable name" and "Variable Value", respectively, with HYPLASHOME and the path name (in full) of the directory ..\HYPLAS_v2.0\bin. 8. Press "OK" on the relevant pop-up windows. 9. The next time the computer is REBOOTED, this variable will be set to the correct path and HYPLAS should be able to find the error messages file ERROR.RUN if required. UNIX/LINUX systems ------------------ In a UNIX/LINUX operating system using a C-shell, for instance, the HYPLASHOME environment variable should be set with the command: setenv HYPLASHOME where here denotes the full path to the directory ../HYPLAS_v2.0/bin. To compile HYPLAS (from directory ../HYPLAS_v2.0/src) with a FORTRAN 77 compiler such as g77, you can use the command: g77 -o ../bin/hyplas hyplas.f */*.f Note that the executable file "hyplas" will be stored in the directory ../HYPLAS_2.0/bin (i.e. the directory set in the HYPLASHOME environment variable). Alternatively, you may use the Makefile provided (with suitable modifications, if needed) to create the HYPLAS executable. IMPORTANT: Before generating a HYPLAS executable, read Sections 2.(a) and 2.(b) below. 2.(a) Memory Requirements ------------------- HYPLAS memory requirements depend on the array dimensioning parameters set in files: ../HYPLAS_v2.0/src/ ELEMENTS.INC GLBDBASE.INC MATERIAL.INC MAXDIM.INC Files ELEMENTS.INC, GLBDBASE.INC and MATERIAL.INC contain parameters which are associated with the currently implemented finite elements and materials. DO NOT MODIFY THEM ! unless you are absolutely sure of what you are doing (only developers coding new elements or new material models/analysis types may need to modify them by changing the existing dimensioning parameters and/or including new parameters). The ONLY dimensioning file that can be safely modified by the average user is the file MAXDIM.INC This file contains the array dimensioning parameters related to the maximum permissible dimension of problems to be analysed by HYPLAS. These parameters include the maximum number of nodes, elements, element groups, etc. If necessary, CHANGE THESE PARAMETERS TO SUIT YOUR PROBLEM SIZE/MEMORY REQUIREMENTS before compiling HYPLAS. 2.(b) Testing a newly compiled executable ----------------------------------- After you have successfully compiled the HYPLAS source code and created an executable file, the next step is to run some tests to verify that HYPLAS is working well. To do this, proceed as follows: The directory ../HYPLAS_v2.0/book_examples/data_files contains a series of data files named .dat of benchmarked examples described in the companion textbook. The corresponding (benchmarked) result files are in the directory ../HYPLAS_v2.0/book_examples/result_files This directory contains a series of result files named .res generated with the current version of HYPLAS on a tested platform. All these files have been named such that their names start with the textbook section number where the corresponding example is described. For instance, files 14_9_2_tresca.dat and 14_9_2_tresca.res refer to a problem described in section 14.9.2 of the textbook, and so on. To check that HYPLAS is working well on your platform, after compiling HYPLAS, run the program HYPLAS for the examples of files .dat and compare the newly generated results .res with their benchmarked counterparts (of the same filename) in the result_files directory. To run an example, execute HYPLAS and use the keyboard to enter the name of the corresponding data file in full (including the extension .dat). To compare the benchmarked .res files against their newly generated you may proceed as follows: 1. On MICROSOFT WINDOWS systems - Here we have successfully used the software "ExamDiff" (the task was made particularly easy by selecting "View" and then the "Show Differences Only" option - this refers to version 1.8 of this software). 2. On UNIX/LINUX systems - Here we use the "diff" command from a shell window (and set the option to ignore blank spaces). A shell script may be used to perform this task automatically (including running HYPLAS and checking for result file differences) for all benchmarked examples provided. IMPORTANT: THE ONLY ACCEPTABLE DIFFERENCES BETWEEN A THE NEWLY GENERATED RESULT FILES AND THEIR BENCHMARKED COUNTERPARTS ARE THE DIMENSIONING PARAMETERS (FROM FILE MAXDIM.INC) USED TO COMPILE THE NEW EXECUTABLE (THESE PARAMETERS ARE PRINTED RIGHT AT THE BEGINNING OF THE RESULT FILES) AND NUMERICAL DIFFERENCES IN RESULTS DUE TO NUMERICAL "ROUNDING-OFF" (THESE ARE VERY SMALL DIFFERENCES THAT DEPEND ON THE PRECISION OF ARITHMETIC OPERATIONS IN THE PLATFORM USED). ALSO NOTE THAT THE EXAMPLES OF THE COMPANION TEXTBOOK DO NOT COVER ALL FEATURES OF HYPLAS. HENCE THIS TEST DOES NOT GUARANTEE THAT EVERYTHING IS WORKING PROPERLY. 3. THE H Y P L A S DIRECTORY TREE ================================ 3.(a) Summary ------- ../ HYPLAS_v2.0/ bin/ book_examples/ data_files/ result_files/ man/ html/ src/ CRYSTAL/ DAMAGE/ DAMAGED_ELASTIC/ DRUCKER_PRAGER/ ELASTIC/ ELEMENTS/ GENERAL/ MATERIALS/ MATHS/ MOHR_COULOMB/ OGDEN/ TRESCA/ VON_MISES/ VON_MISES_MIXED/ 3.(b) Description ----------- The HYPLAS program directory tree is organised as follows: ../HYPLAS_v2.0/ (this directory) This is the HYPLAS root directory, where the HYPLAS directory tree starts. ../HYPLAS_v2.0/bin/ This directory contains the file ERROR.RUN where most HYPLAS error/warning messages are. IMPORTANT: the environment variable HYPLASHOME should be set to this directory. Otherwise, HYPLAS will not find its error/warning messages when required. We also recommend that the EXECUTABLE of HYPLAS be stored in this directory. ../HYPLAS_v2.0/book_examples/ This directory has the following subdirectories: ../HYPLAS_v2.0/book_examples/data_files ../HYPLAS_v2.0/book_examples/result_files Refer to Section 2.(b) above for further details. ../HYPLAS_v2.0/man/ This is the HYPLAS documentation/manuals directory. It contains the following files: input_man.txt - A concise input data manual for HYPLAS in ASCII format; hyplas_calltree.txt - Contains a flowgraph (shows the call tree) of HYPLAS in ASCII-format. Note: calls to function subprograms are not included in this flowgraph; and the subdirectory: ../HYPLAS_v2.0/man/html This directory contains the hypertext (HTML) format Fortran source code and of manual pages of the entire HYPLAS program. Manual pages with descriptions of each function/subprogram including their argument list are linked to their corresponding HTML-format source code. This allows the user the navigate through the HYPLAS source code using a web browser. To start at the main program, use your web browser to open the file hyplas.html. This facility should be helpful to those trying to understand the flow of program HYPLAS. ../HYPLAS_v2.0/src/ This directory (and its subdirectories) contains the Fortran source code of HYPLAS. The files containing the sources are named following the standard practice: .f where is the name of the FORTRAN procedure (subroutine, function subprogram, etc.) whose source code is in file .f. The source code of the HYPLAS main program is in file hyplas.f and the HYPLAS database (COMMON blocks, array dimensioning parameters and other global parameters) is coded in the "include files" ELEMENTS.INC GLDBASE.INC MATERIAL.INC MAXDIM.INC in this directory. In addition, this directory contains a file named "Makefile" (UNIX-LINUX Release only) which may be used for compiling and linking HYPLAS in UNIX/LINUX systems. The subdirectories of ../HYPLAS_v2.0/src are as follows: ../HYPLAS_v2.0/src/CRYSTAL Contains the source code of all procedures related to the finite strain single crystal plasticity model implemented in HYPLAS. ../HYPLAS_v2.0/src/DAMAGE Source files of the procedures related to the Lemaitre ductile damage model implementation. ../HYPLAS_v2.0/src/DAMAGED_ELASTIC Source files of the procedures related to the damaged elasticity model with crack closure effect. ../HYPLAS_v2.0/src/DRUCKER_PRAGER Source files of the procedures related to the implemented Drucker-Prager plasticity model. ../HYPLAS_v2.0/src/ELASTIC Source files of the procedures related to the linear elasticity model (Hencky model under large strains) implemented. ../HYPLAS_v2.0/src/ELEMENTS Source files of the element interfaces and element-related procedures. ../HYPLAS_v2.0/src/GENERAL Source files of general procedures. ../HYPLAS_v2.0/src/MATERIALS Source files of the material interfaces. ../HYPLAS_v2.0/src/MATHS Source files of the mathematical procedures. ../HYPLAS_v2.0/src/MOHR_COULOMB Source files of the procedures related to the implemented Mohr-Coulomb plasticity model. ../HYPLAS_v2.0/src/OGDEN Source files of the procedures related to the implemented Ogden hyperelasticity model. ../HYPLAS_v2.0/src/TRESCA Source files of the procedures related to the implemented Tresca plasticity model. ../HYPLAS_v2.0/src/VON_MISES Source files of the procedures related to the implemented von Mises plasticity model with isotropic hardening. ../HYPLAS_v2.0/src/VON_MISES_MIXED Source files of the procedures related to the implemented von Mises plasticity model with mixed isotropic/kinematic hardening. 4. CROSS-REFERENCING BETWEEN THE SOURCE CODE AND THE TEXTBOOK ========================================================== Many references are made in the textbook to various subprograms of HYPLAS. These are usually made when a particular procedure described in the text is implemented in the program. The reader should refer to the textbook index. Also, a substantial number of comment lines have been added to the source code of HYPLAS with reference to sections, figures, boxes, etc of the textbook related to the part of the code in question. Such references are usually displayed after the word "REFERENCE:" (in capitals) on commented lines. Searching for this word will take you to the line of code where the particular routine has a reference to the textbook. NOTE: Occasional references to other textbooks/journal papers are also made following the word "REFERENCE:" on commented lines. 5. HYPLAS ERROR MESSAGING ====================== Most error/warning messages issued by HYPLAS are in the ASCII-format file ERROR.RUN (kept in the HYPLASHOME directory - ../HYPLAS_v2.0/bin). All such error/warning messages have an identification code (e.g. ED0015) which is printed both to the standard output (this is usually the computer screen) and to the relevant results file. If you wish to find where in the source code a particular message is being issued, then perform a search for the corresponding message identification code in the entire source code of HYPLAS. 6. FURTHER REMARKS ON HYPLAS ========================= 6.(a) Program efficiency THIS SECTION IS OF INTEREST ONLY TO THOSE WANTING TO MAKE HYPLAS RUN FASTER. It is particularly stressed in the textbook that this program has not been designed having efficiency in mind (refer to Section 5.1.2 of the textbook). Its structure has been designed mainly to illustrate in a relatively clear manner the computer implementation of the techniques and algorithms described in the text, with a particular view to the implementation of solid constitutive models and finite elements. For those who are especially interested in the speed of the code, there are a few tips that could help in this direction. Unfortunately, these involve modifications to the source code which is probably most appropriate to readers with a good level of experience in finite element programming. To those with this particular interest, we can suggest the following: (i) The use of faster linear solvers This is probably the change that would result in a greater gain in efficiency. The Frontal Method adopted in subroutine FRONT (file ../HYPLAS_v2.0/src/GENERAL/front.f) has been designed originally to save memory (back in the days when computer memory was severely limited). There are currently a vast number of methodologies which focus on speeding up the linear solution, in addition to reducing memory storage requirements (which is a particularly important issue in the solution of large scale problems). Some of these are extensions/refinements of the original Frontal solver. We remark that a number of such procedures (with their respective source codes) are available (conditions may apply) from the LAPACK (Linear Algebra PACKage - http://www.netlib.org/lapack) repository or from the HSL Library (http://www.cse.cse.scitech.ac.uk/nag/hsl). For the reader interested in gaining speed, we would recommend the replacement of the existing solver of FRONT by a faster one. We remark though that this is a substantial programming task. Another aspect here is the fact that computing times in FRONT are directly linked to the frontwidth of the system which, in the present version of HYPLAS is fixed and depends, for a given mesh, on how the degrees of freedom are numbered (node numbering). The incorporation of a frontwidth optimiser (which re-numbers the degrees of freedom in order to minimise the frontwidth) in FRONT could produce some good savings in computing times. Such savings become particularly noticeable in larger problems where the original node numbering produces an excessively large frontwidth. (ii) Material-specific computations The issues pointed out here affect only the computing times for specific material models and are expected to have a much lower impact in overall speed than the linear solver issue discussed above. Some of the material model-specific computations carried out in HYPLAS could be made a bit faster. For example, for isotropic models whose stress update is carried out in the principal stress space (such as the Tresca and Mohr-Coulomb models - see routines SUTR and SUMC, files ../HYPLAS_v2.0/src/TRESCA/sutr.f and ../HYPLAS_v2.0/MOHR_COULOMB/sumc.f, respectively) the spectral decomposition of the stress in carried out in the state update update routine and then repeated in the corresponding routine for computation of the consistent tangent operator (refer to files ../HYPLAS_v2.0/src/TRESCA/cttr.f and ../HYPLAS_v2.0/src/MOHR_COULOMB/ctmc.f, respectively, for the Tresca and Mohr-Coulomb plasticity models). Some savings in computing time can be achieved here by storing the stress eigenprojection tensors (these can be stored as state variables) during the execution of the state updating and then retrieving them later for use in the computation of the consistent tangent operator. This change can be incorporated to the code relatively easily. The computation of the exponential map and is derivative for the single crystal plasticity model (routines EXPMAP, file ../HYPLAS_v2.0/src/CRYSTAL/expmap.f and DEXPMP, file ../HYPLAS_v2.0/src/CRYSTAL/dexpmp.f) is carried out in three dimensions (these routines have been adapted from an earlier three-dimensional code). To improve efficiency, these can be adapted to work only in two-dimensional problems by removing the unnecessary operations related to the third dimension. 6.(b) Output of nodal averaged values The reader should be aware that the way in which nodal averaged values of stresses and other variables are calculated in HYPLAS is very basic (and rudimentary). This feature of the program is made available only to help those interested in producing contour plots, etc from results presented in HYPLAS result files and should be useful in many circumstances of interest. This facility has in fact been used in producing many of the figures presented in the textbook. But note, for example, that the values of incremental plastic multipliers for plasticity models may take (inadmissible) negative values when extrapolated from Gauss-point to nodes and averaged. We remark that more sophisticated and refined techniques of transferring Gauss point values of variables to nodal points and obtaining the corresponding smoothed field are available in the current literature. These fall outside the scope of the companion textbook of HYPLAS. Platform: |
Size: 11008084 |
Author:gtcewli3 |
Hits:
Description: 不同算法对时间序列进行奇异谱分解的源程序代码，matlab2006a平台调试通过-different algorithms on time series singular spectral decomposition of the source code, matlab2006a through debugging platform Platform: |
Size: 6144 |
Author:mapleyu |
Hits:
Description: 旋转不变子空间类算法是空间谱估计中的典型算法，它也是特征分解类算法-Rotation-invariant subspace type algorithm is the space of a typical spectral estimation algorithm, it is also eigen-decomposition algorithm Platform: |
Size: 1024 |
Author:马连达 |
Hits:
Description: 何振亚的现代谱估计.本书全面系统地论述了现代谱估计技术中的各种理论与方法。全书共九章，内容包括；纯连续谱沽计的AR与ARMA模型参量法、纯离散谱与混合谱估汁的正弦组合与阻尼复指数模型参量法、非参量法谱估计的最小方差法与奇异值/特征值分解处理法、应用信息论的墒谱估汁法、多谱(高阶谱)估计、二维潜估计与多维阵列谱估计。本书取材广泛，内容新颖，理论联系实际，系统性强。
-He Zhenya Modern spectral estimation. This book comprehensively and systematically discusses the modern spectral estimation techniques in a variety of theories and methods. Book a total of nine chapters, including pure continuous spectrum of sell ARMA model with AR parameter method, pure discrete spectrum and mixed spectrum estimation of sinusoidal juice combination of complex exponential model with the damping parameter method, non-parameter method of minimum variance spectral estimation method with the singular value/eigenvalue decomposition treatment method, the application of information theory the entropy spectrum estimation juice method, multi-spectral (HOS) estimates that two-dimensional latent estimation and multi-dimensional spectral estimation array. The book based on a wide range of content of a novel theory with practice, systemic strong. Platform: |
Size: 6134784 |
Author:songzy41 |
Hits:
Description: 该程序提供了多种参数的双指数脉冲函数的波形及其频谱分解图，要注意的是作图时坐标比例尺的调整。程序中给出的是1.2/50μs和0.7/60ns两种波形可直接使用的坐标比例尺。
在MATLAB5.0下就能工作。
-The program provides a wide range of parameters of double-exponential pulse function waveform and its spectral decomposition diagram, it should be noted that when mapping coordinates scale adjustments. Procedure is given in 1.2/50μs and 0.7/60ns two waveforms can be directly used the coordinates of scale. Will be able to work under in MATLAB5.0. Platform: |
Size: 1024 |
Author:JiYongLiang |
Hits:
Description: 语音降噪。从Codec AD50采集话筒语音，通过DSP TMS320vc5402处理，在送到AD50输出降噪后语音，涉及加汉宁窗，切比雪夫滤波器，快速傅立叶变换和反FFT，有声无声判断谱分解，谱合成等功能-Voice noise. Codec AD50 collected from the microphone voice, through the DSP TMS320vc5402 treatment, in AD50 to the output noise after the voice, involve Hanning windows, Chebyshev filter, Fast Fourier Transform and anti-FFT, audio silent judge spectral decomposition, spectral synthesis and other functions Platform: |
Size: 44032 |
Author:黄胜华 |
Hits:
Description: 现代谱估计的matlab的程序。用的是奇异值分解的总体最小二乘法。-Modern Spectral Estimation matlab procedures. Using a singular value decomposition of the total least squares method. Platform: |
Size: 31744 |
Author:yangqiuchen |
Hits:
Description: 文章介绍了MPD得基本原理，对于谱分解技术研究具有一定的指导意义-This paper introduces the basic principles of a MPD for the spectral decomposition technology research has a certain significance Platform: |
Size: 2098176 |
Author:ljc |
Hits:
Description: 一种自适应chirplet分解的快速算法
针对信号自适应chirplet分解未知参数多、实现起来比较困难的特点，文献[1]提出了一种新的chirplet分解快速算法。该算法利用计算信号的二次相位函数，得到其能量分布集中于信号的调频率曲线上的结论，此时通过谱峰检测可同时获得chirplet调频率、时间中心和幅度的估计；然后通过解线性调频技术获得其初始频率和宽度的估计，仿真结果验证了本文算法的有效性。
-Chirplet an adaptive decomposition algorithm for rapid adaptive signal decomposition Chirplet many unknown parameters, is more difficult to achieve the characteristics of the literature [1] proposed a new fast algorithm Chirplet decomposition. The algorithm calculating the signal function of the second phase, its energy distribution to be concentrated in the signal modulation frequency curve on the conclusions, this time through the spectral peak detection can be obtained simultaneously Chirplet tune the frequency, timing and magnitude of the estimated center then through the solution line technology of FM frequency and the width of the initial estimates, the simulation results verify the effectiveness of the algorithm. Platform: |
Size: 80896 |
Author:nanzhewan |
Hits:
Description: 实现EMD分解，并对原始信号进行hilbert谱分析，通俗易懂，操作方便-The realization of EMD decomposition, and the original signal hilbert spectral analysis, user-friendly, easy to operate Platform: |
Size: 81920 |
Author:joice5 |
Hits:
Description: MPD is a decomposition spectral method for seismic data and also for signal analysis too. It works based on atoms, book, and dictionary. We start from WPP Pursuit, incorporating another four files in a one worksheet on MATLAB. Platform: |
Size: 8192 |
Author:humbang |
Hits:
Description: 基于奇异值分解的人脸识别方法
梁毅雄 龚卫国 潘英俊 李伟红 刘嘉敏 张红梅
提出了一种将傅里叶变换和奇异值分解相结合的人脸自动识别方法.首先对人脸图像进行傅里叶变换,得到其具有位移不变特性的振幅谱表征.其次,从所有训练图像样本的振幅谱表征中给定标准脸并对其进行奇异值分解,求出标准特征矩阵,再将人脸的振幅谱表征投影到标准特征矩阵后得到的投影系数作为该人脸的模式特征.然后,对经典的最近邻分类器算法进行了改进,并采用模式特征之间的欧式距离作为相似性度量,从而完成对未知人脸的识别.采用ORL (Olivetti Research Laboratory)人脸库对本文提出的人脸识别方法进行验证,获得了100.00 的识别率.实验结果表明,本方法优于现有的基于奇异值分解的人脸识别方法,且对表情、姿态变换等具有一定的鲁棒性.
-Face recognition based on singular value decomposition method
Deliberate simultaneously Gong Weiguo Li Wei Hung Stephen Lau, Hong-Mei Zhang Ying-Jun Pan
Paper, a Fourier transform and singular value decomposition of the combination of automatic face recognition. First of all, the face image by Fourier transformation, it has the same characteristics of the displacement amplitude spectra. Secondly, all training The amplitude spectrum of the sample images given in standard face representation and its singular value decomposition, find the standard characteristic matrix, then the amplitude of spectral characterization of human faces projected onto the standard characteristic matrix of projection coefficients obtained as the face of the model features . Then, the classical nearest neighbor classifier is improved, and the use of Euclidean distance between pattern features as the similarity measure, thus completing the identification of unknown human faces. using ORL (Olivetti Research La Platform: |
Size: 58368 |
Author:houhj |
Hits:
Description: 优秀论文及配套源码。Hilbert-Huang变换(HHT)是一种新的非平稳信号处理技术，该方法由经验模态 分解(EMD)与Hilbert谱分析两部分组成。任意的非平稳信号首先经过EMD方法处理后被分解为一系列具有不同特征尺度的数据序列，每一个序列称为一个固有模态函数(IMF)，然后对每个IMF分量进行Hilbert谱分析得到相应分量的Hilbert谱，汇总所有Hilbert谱就得到了原信号的谱图。该方法从本质上讲是对非平稳信号进行平稳化处理，将信号中真实存在的不同尺度波动或趋势逐级分解出来，最终用瞬时频率和能量来表征原信号的频率含量。
本文研究了基于HHT的暂态电能质量扰动检测方法，介绍了HHT的基本原理和利用HHT检测电能质量多扰动信号的实现方法。仿真试验表明该方法可以实时检测扰动的起止时刻，持续时间和扰动幅度，适用于电能质量多扰动的监测和辨识系统。
-Hilbert-Huang Transform(HHT) is a new non-stationary signal processing technology，including the Empirical Mode Decomposition(EMD) and Hilbert spectral analysis．In this method，arbitrary non-stationary signals was divided into a series data sequence with different characteristic scales after the first EMD treatment，each sequence is called an intrinsic mode function(IMF)，and then each IMF component gained corresponding the Hilbert spectrum though Hilbert spectrum analysis，a summary of all Hilbert spectrum assembled to the spectra of the original signal．In essence，this method is a stable processing for non-stationary signal，which decomposed the different fluctuations or trends in the signal in rum，and represented the frequency content of the original signal using the final instantaneous frequency and energy．
This paper study a method to detect power quality disturbances based on the HHT，in which the principle and the realization of the HHT and the realization of power quality disturbance Platform: |
Size: 321536 |
Author:NBB |
Hits:
Description: 本书介绍了图像去模糊化的一些基本的知识，并且主要针对svd和spectral decomposition的去模糊算法进行了介绍，以及对于这些方法的一些提高计算速度的算法进行了详细是介绍-This ebook intruduces some basic knowledge of deblurring, especially the methods of SVD and Spectral Decomposition, and some fast algorithms for the two methods. Platform: |
Size: 6670336 |
Author:Unique |
Hits:
Description: Matlab的关于EMD的几篇文章
谱和主成份分析用于EMD分解研究；基于EMD方法的混沌时间序列预测-Matlab several articles on EMD and principal component analysis for spectral decomposition of EMD EMD-based method of chaotic time series prediction Platform: |
Size: 254976 |
Author:ifengdoo |
Hits:
Description: 用傅里叶变换实现地震资料频谱分解的源程序，本程序是基于fortran90编写-Seismic data using Fourier transform spectral decomposition of the source, this procedure is based on the fortran90 write. Platform: |
Size: 2048 |
Author:徐维 |
Hits:
Search - spectral decomposition