最专业的信号处理分析软件之一

包括以下:
傅立叶变换 FFT
自动回归 AutoRegressive
移动平均回归 Moving Average
自回归移动平均模型分析 ARMA
复数指数模型 Complex exponential modeling
最小变异数 Minimum variance methods
特征分析 Eigen analysis frequency estimation
小波分析 Wavelets



More Information:
http://www.cnTech.com.cn/

中仿科技有限公司是专业从事信息化软件以及咨询服务的提供商，是该SYSTAT系统软件（包括SYSTAT软件、SigmaPlot软件、SigmaScan软件、SigmaStat软件、TableCurve2D/3D软件、PeakFit软件、AutoSignal软件）在中国地区的独家代理商，是中国地区唯一授权销售SYSTAT系统软件的公司。如果您需要咨询我们的产品信息、软件升级信息，请不要犹豫与我们进行联系。同时欢迎访问SYSTAT美国网站：http://www.systat.com/ ，您可以获得更详尽的信息。

AutoSignal is the first and only program that completely automates the process of analyzing signals. It gives researchers the power to rapidly find components of complex signals that normally require extensive programming and mathematical routines. It takes full advantage of its graphical user intuitive interface to simplify every aspect of operation, from data import to output of results.

AutoRegressive linear models offer robust models that can quickly handle smaller data sets that FFT cannot accurately analyse. With AutoSignal, you can also recover signal components based on power - the component may be sinusoidal, a square wave, a sawtooth or anharmonic pattern. It lets you see a complete picture of the frequency space using the library of six Fourier Spectrum methods with total flexibility. It gives you a choice of three adjustable mother wavelets: Morlet, Paul and Gaussian Derivative - in both real and complex forms to optimise localisation results.

Scientists and engineers can perform complex signal analysis without programming by selecting menu items that determine how the computer will analyze and present data. Selecting a processing step or algorithm causes the software to provide further menus or windows so users can tailor performance to their needs. After the user makes a processing selection, the software immediately presents processing results in 2-D or 3-D views

Built-in spectral analysis procedures include :

FFT
AutoRegressive
Moving Average
ARMA
Complex exponential modeling
Minimum variance methods
Eigen analysis frequency estimation and Wavelets
AutoSignal consists of many important leading edge signal analysis, filtration, prediction, and noise removal technologies. Instead of combining these into a single interface, each of these has been assembled into separate mini-applications. This approach allows each procedure to be custom tailored and streamlined for ease of use. The program consists of four basic sections:

Data Import, Generation, and Basic Modification (File, Edit, Data menus; 'Main' toolbar)
Time Domain Analysis Procedures (Time menu, 'Time' toolbar)
Frequency and Time-Frequency Spectral Analysis Procedures (Spectral menu, 'Spectral' toolbar)
Advanced Processing Procedures in Fourier, Wavelet, and Eigendecomposition Domains (Process menu,'Process' toolbar)
Disciplines where AutoSignal can be applied :
Communications - signal identification and analysis
Signal interference monitoring
Control system analysis
Audio system analysis
Voice recognition and speech processing
Signature analysis
Vibration research and analysis
Acoustic research and analysis
Radar signal analysis
Analogue circuit testing
Signal detection
Oceanography
Geological Sciences
Seismology and Marine Geophysics
Meteorology
Astrophysics
Electrical Engineering
Mechanical Engineering
Civil Engineering
Aerospace Engineering
Defence research
Features
Perform complex signal analysis with a mouse click
Quickly locate your signal components
Precisely estimate with advanced parametric modelling
Easily smooth and process your signals
Graphically review signal analysis results
Identify frequency and power with Fourier Spectrum analysis
Effortlessly analyze non-stationary data with wavelets
Isolate components by signal strength using eigendecomposition
Save precious research time with the production facility
System Requirements
Windows 95, 98 and NT
Pentium or clone and above
32 MB RAM minimum (64 MB RAM for wavelet and production facility recommended)
25 MB hard disk space
SVGA and above
TOPICS



Perform complex signal analysis with a mouse click
Quickly locate your signal components
Precisely estimate with advanced parametric modelling
Easily smooth and process your signals
Graphically review signal analysis results
Identify frequency and power with Fourier Spectrum analysis
Effortlessly analyze non-stationary data with wavelets
Isolate components by signal strength using eigendecomposition
Save precious research time with the production facility



Perform complex signal analysis with a mouse click - no programming required !

AutoSignalTM is the first and only program that completely automates the process of analyzing signals. Save precious time by eliminating the programming time normally required for performing sophisticated signal analysis. AutoSignal takes full advantage of its graphical user intuitive interface to simplify every aspect of operation, from data import to output of results. Choose your analysis techniques from the menu or toolbar. Select the algorithm and options from the interface. You get immediate visual feedback with 2D or 3D graphs of your signal analysis plus numeric summaries for reports.




Click to view larger image

Quickly locate your signal components

AutoSignal gives researchers the power to rapidly find components of complex signals that normally require extensive programming and mathematical routines. AutoSignal provides a vast array of spectral analysis procedures to help you make intelligent conclusions for any application. Built-in spectral analysis procedures include:


FFT
AutoRegressive
Moving Average
ARMA
Complex exponential modeling
Minimum variance methods
Eigen analysis frequency estimation and Wavelets


Precisely estimate with advanced parametric modeling


With AutoSignal, you get state-of-the-art parametric nonlinear modeling for sinusoid and damped sinusoid models. Non-linear optimization is also available as an independent procedure, or as an adjunct to each of the spectral algorithms. It includes robust maximum-likelihood optimizations as well as automatic parameter constraints. AutoRegressive linear models offer robust models that can quickly handle smaller data sets that FFT cannot accurately analyse.




Click to view larger image

Easily smooth and process your signals

Only AutoSignal offers so many different user-friendly methods to manipulate signal data. You can inspect your data stream in the Fourier domain and zero higher frequency points - and see your results immediately in the time domain. This smoothing technique allows for superb noise reduction while maintaining the integrity of the original data stream. AutoSignal also includes eigendecomposition, wavelet, Savitzky-Golay, Loess and detrending for smoothing and denoising. Isolate components and detect signals with powerful filtering and reconstruction techniques with Fourier, eigendecomposition and wavelet methods. For instance, isolate components that appear and disappear with wavelet filtering and reconstruction. Recover the true signal that would have been measured using an ideal sensing system with Gaussian and exponential deconvolution.


Graphically review signal analysis results

As a powerful visualization tool, AutoSignal automatically plots your peaks, contours or 3D surfaces - so you don't have to perform additional steps to see your results. Change any algorithm or analysis option on the fly through the user interface and see instant results. Isolate components of a signal graphically using eigen decomposition to display and select eigen components in order to find very low frequency oscillatory components or identify paired eigen modes producing a specific oscillation. Analyze your results with residual and root plots, and show statistical significance and probability limits on your output graphs. Clearly present your results with control over titles, fonts, colors, points, scaling, axis scale, labels, grid and plot types.


Identify frequency and power with Fourier Spectrum analysis


AutoSignal lets you see a complete picture of the frequency space using the library of six Fourier Spectrum methods with total flexibility. Solve the leakage problem found with standard FFT by using one of the 30 included data tapering windows. You can even make comparisons of performance of various data tapering windows in a single spectral graph. AutoSignal gives you access to the latest methodologies with techniques such as FFT Multi-taper Spectrum analysis to help you better characterise the power in each signal. Easily handle your unevenly sampled data with Lomb-Scargle Fourier domain analysis with techniques that were originally developed by astrophysicists.


Click to view larger image


Effortlessly analyze non-stationary data with wavelets

Simultaneously find the time and frequency localisation components of a non-stationary periodic signal with Continuous Wavelet Spectrum analysis techniques. AutoSignal gives you a choice of three adjustable mother wavelets: Morlet, Paul and Gaussian Derivative - in both real and complex forms to optimise localisation results. You can also perform power analysis in either the time or the frequency range with specialised in-depth analysis techniques to evaluate the signal.


Isolate components by signal strength using eigendecomposition


In addition to FFT and wavelet spectral analysis techniques, you can select from linear and non-linear methods that are right for your application. The eigendecomposition procedures enable you to visually select eigenmodes for signal-noise separation or component isolation. With AutoSignal, you can also recover signal components based on power - the component may be sinusoidal, a square wave, a sawtooth or anharmonic pattern. You can confirm the presence of white noise or isolate red noise by reconstructing only the noise eigenmodes.


Click to view larger image

Save precious research time with the production facility


What once took hours now takes seconds - with only a few mouse clicks. It's so easy - even novice users can learn how to use AutoSignal in no time. Every procedure is automated. For even more muscle, streamline your work with the production facility to automate batch analysis and reporting. With an easy-to-use dialog, set up your batch import and export options. Link directly to your hardware to analyse and report on the fly. Already have your data in Microsoft® Excel? No problem. Process up to 255 Excel worksheets at once. Create RTF reports with numerical summaries that include publication-quality graphs or export the data to a new Excel workbook. With AutoSignal, it's so simple!

GENERAL FEATURES

Interface

Full 32-bit performance
Toolbars and menu-driven for all functionality
No programming of procedures, algorithms or graphing
Advanced on-line help system
Non-Parametric Spectral Analysis

AR spectral methods offer accurate frequency estimation with short data records
AR spectrum methods: AR spectrum, AR with order exploration, AR with algorithm comparison *
14 AR algorithms: autocorrelation method, maximum entropy method (Burg), least-squares normal equations, least-squares covariance models and modified covariance models, singular value decomposition methods

Model order selection and order exploration
Moving average spectrum
ARMA spectrum
Prony Spectrum offers fitting of damped sine and damped exponential that occur in multi-component exponential decays
Minimum Variance Spectrum
Eigen Analysis Spectrum

Eigen Analysis Spectrum provides accurate and robust spectral procedures for estimating harmonic frequencies
Provides excellent signal-noise separation
Graphically select signal and noise sub-space; also available in certain parametric procedures
Data Processing

Non-Linear Optimization offers parametric refinement of spectral estimates: least-squares, maximum likelihood
Fourier Interpolation
Fourier Upsampling
Parametric Interpolation and Prediction
Graphically inspect the autocorrelation series
Detrend: Constant, Linear, Quadratic, Cubic, Logarithmic, Exponential, Power, Hyperbolic
Difference the data with adjustable order and lag, compute various cumulatives, and normalize for unit area, unit power, unit standard deviation, and zero mean
Add or subtract a reference signal
Compare imported reference signals
Gaussian deconvolution or exponential deconvolution to remove instrument response smearing *
Find long-term 'memory effects' in flat frequency response signals with Fractal Dimension option
Numeric Review

Full component numeric summary report include: procedure, algorithm, listing of interpolated spectral peaks, frequency analysis and linear least-squares fit summary
List data offers extended data summary with results from each of the procedure such as frequency, magnitude, phase and power spectral density
Goodness of fit statistics: r2, degrees of freedom adjusted r2, fit standard error, F-statistic
Evaluation option with automated table generation, includes function, derivatives, roots and cumulative volume; X values can be generated or imported from file
Output and Export

Publication-quality printed graphs
Image formats include bitmaps, metafiles, enhanced metafiles and device-independent bitmaps
File formats include ASCII, Excel, Lotus, SYSTAT, SPSS
Export numerical summaries and graphs to MS Word RTF documents
Data Input

Up to 65,536 points in data table
Over 65.4 million points can be filtered into table using decimation import filter
ASCII (Single, X-Y, and Multi-column)
Excel (Excel 97, Excel 95, v5, v4, v3)
Lotus 123 (WK4, WK3, WK1)
Quattro Pro (WB2, WB1)
SigmaPlot (JPG, SPW, SP5)
SPSS (SAV v7.5, v8 and v9)
SYSTAT (SYD v8)
Waveform (WAV MS PCM 8, 16, 32 bit)
DIF (Single, X-Y, and Multi-column)
dBase (DBF III+, IV)
Import Preview graphs prospective data
Separate append options that automatically averages replicates
Fourier Spectral Analysis

Procedures: Fourier Spectrum, Fourier Spectrum with Data Window, Fourier Spectrum with Data Window Comparison *, Fourier Spectrum of Segmented Data, Fourier Multitaper Spectra, Fourier Spectrum of Unevenly Sampled Data (Lomb-Scargle)

Transforms: FFT Radix2, Prime Facto, Mixed Radix, Chirp-Z, Best Exact-N
Zero pad
30 tapering windows:
Fixed: none, Welch, Bisquare, Bartlett, cs2 Hanning, Tukey-Hanning, cs2 Hamming, Bartlett Mod, cs3 Nuttall C3, cs3 Blackman, cs3 Blackman-Harris 3, cs3 Nuttall C1, cs3 Blackman Exact, cs3 Blackman-Harris min, cs3 Nuttall min, cs4 Nuttall C5, cs4 Blackman-Harris 4, cs4 Nuttall C3, cs4 Nuttall C1, cs4 Blackman-Harris min, cs4 Nuttall min
Adjustable: Beta, csx max Roll-off, Kaiser-Bessel, VanderMaas, Chebyshev, Chebyshev Appr, Slepian DPSS, Gaussian, Tapered-Cosine

Compare up to 4 tapering windows simultaneously
Measure data window properties: mainlobe, sidelobe, roll-off
Time-Frequency Spectral Analysis

Short-Time Fourier Transform Spectrum uses a series of segmented and overlapped FFTs to find Fourier spectral information for non-stationary data
Continuous Wavelet Spectrum multi-resolution time-frequency techniques: 3D surface, contour, power integration across time or frequency
Wavelet spectra can be generated with up to 100 linear or logarithmic frequencies
Adjustable mother wavelets: Morlet, Paul, Gaussian Derivative
Zero padding available
Full critical significance limits available as 3D gradients
Graphical rendering of cone of influence
Automated power analysis by integrating interpolated wavelet spectrum surface
Filtering and Reconstruction

Fourier Smoothing and Denoising: frequency or signal threshold filtration
Eigen decomposition Smoothing and Denoising: signal strength threshold filtration
Wavelet Smoothing and Denoising: thresholds in the time-frequency domain for non-stationary data.
Fourier Filtering and Reconstruction: Fourier domain filtering and component isolation procedure
Eigen decomposition Filtering and Reconstruction: isolates individual oscillatory components in signals
Wavelet Filtering and Reconstruction: isolates in the time-frequency domain
Savitzky-Golay Smoothing filter: includes smoothing to 1st through 4th derivative
Spline Estimations: cubic, cubic constrained, smoothing cubic, B-spline, B-Spline Fix knots, B-spline Optimal knots, NURBS
Adjustable order Loess with tricube and Gaussian weighting
Significance Levels

Unique Peak-based critical limit levels to ascertain the significance of the spectral components
Critical limits plotted are: 50%, 90%, 95%, 99%, 99.9% (uses color gradients for wavelets)
Peak-type critical limits are based on Monte Carlo trials with algorithms exactly as implemented
Can specify AR-1 red-noise background spectrum option
Production Facility *

Batch process large numbers of data sets in an unattended procedure
Import up to 64,515 data sets from Excel 95/97 with up to 255 worksheets and multiple columns per worksheet
Acquire data using simple DLL interface
Automatically export numerical summaries and/or graphs to MS Word RTF documents
Automatically export results of analysis to an Excel 95/97 file