IEST RD-DTE046 1-2011 Terms Commonly Used in the Digital Analysis of Dynamic Data.pdf

1、 Institute of Environmental Sciences and Technology Design, Test, and Evaluation Division IEST-RD-DTE046.1 Reference Document 46.1 Terms Commonly Used in the Digital Analysis of Dynamic Data Arlington Place One 2340 S. Arlington Heights Road, Suite 100 Arlington Heights, IL 60005-4516 Phone: (847) 9

2、81-0100 Fax: (847) 981-4130 E-mail: informationiest.org Web: www.iest.org 2 IEST 2011 All rights reserved Institute of Environmental Sciences and Technology IEST-RP-DTE046.1 This Reference Document is published by the INSTITUTE OF ENVIRONMENTAL SCIENCES AND TECHNOLOGY to ad-vance the technical and e

3、ngineering sciences. Its use is entirely voluntary, and determination of its applicability and suitability for any particular use is solely the responsibility of the user. Use of this Reference Document does not imply any warranty or endorsement by IEST. This Reference Document was prepared by Worki

4、ng Group 019 of the IEST Design, Test, and Evaluation Division and is under the jurisdiction of that division. Copyright 2011 by the INSTITUTE OF ENVIRONMENTAL SCIENCES AND TECHNOLOGY First printing, April 2011 ISBN 978-0-9841330-5-5 PROPOSAL FOR IMPROVEMENT: The Working Groups of the Institute of E

5、nvironmental Sciences and Tech-nology are continually working on improvements to their Recommended Practices and Reference Documents. Suggestions from users of these documents are welcome. If you have a suggestion regarding this document, please use the online Proposal for Improvement form found on

6、the IEST website at www.iest.org. INSTITUTE OF ENVIRONMENTAL SCIENCES AND TECHNOLOGY Arlington Place One 2340 S. Arlington Heights Road, Suite 100 Arlington Heights, IL 60005-4516 Phone: (847) 981-0100 Fax: (847) 981-4130 E-mail: informationiest.org Web: www.iest.org IEST 2011 All rights reserved In

7、stitute of Environmental Sciences and Technology IEST-RD-DTE046.1 3 Institute of Environmental Sciences and Technology Design, Test, and Evaluation Division Reference Document 046.1 Terms Commonly Used in the Digital Analysis of Dynamic Data 1 SCOPE AND LIMITATIONS The purpose of this Reference Docu

8、ment (RD) is to define, clarify, and disseminate terminology appropri-ate for digital analysis of dynamic data as used in envi-ronmental testing. This list is in no manner constrained by any standards publications, technical organizations, or societies. However a reasonable effort has been made to k

9、eep the definitions consistent with these standards. This document is intended for use by those responsible for procuring, manufacturing, and using computer-controlled environmental test systems, as well as those responsible for educating others in the field of envi-ronmental test control. 2 INTRODU

10、CTION This Reference Document (RD) contains certain terms commonly used in digital analysis of dynamic data generated when measuring in-service conditions of mechanical and electrical apparatus. This RD also covers certain conditions encountered during qualifica-tion testing of mechanical systems an

11、d components. Recognizing that there are many ways in which dy-namic data signals are generated, recorded, processed, and displayed, wherever a choice exists we have em-phasized processing of recorded data by digital means. This document is intended to assist those involved in recording, analyzing,

12、displaying, or interpreting data relating to the dynamic properties of mechanical sys-tems. Such data may include strain, vibration, sound pressure variations, and system electrical characteristics. In running text, boldface type is used to indicate termi-nology that is defined in this RD. 3 TERMS A

13、ND DEFINITIONS accelerance See inertance, mechanical impedance. acceleration SRS See shock response spectrum. accuracy The degree of conformity of a measure to a model, standard, or true value. Usually expressed as a per-centage of specified value. Should consider other dependent factors such as tem

14、perature, frequency, relative value (full scale vs absolute), drift rates, and range. aliasing When the sample rate f is less than 2 times the high-est frequency in the data, the frequency is ambi-guously represented. The frequencies above f/2 will be folded back into the lower frequencies to produc

15、e erroneous results. amplitude density function See probability density function. amplitude distribution function See probability distribution function. analog-to-digital converter (ADC) An electronic device used for converting analog sig-nals to digital form, either binary code or binary-coded-deci

16、mal. An ADC is usually used in conjunc-tion with a multiplexer that selects a particular analog signal and with a sample-and-hold amplifier that stores an instantaneous analog voltage. The ADC digitizes (see digitize) that voltage into N discrete levels, where N = 2 and n is the number of bits of re

17、solution in the device. For dynamic waveforms, the sampling frequency must be high enough to avoid aliasing errors. analysis range/bandwidth See frequency range. 4 IEST 2011 All rights reserved Institute of Environmental Sciences and Technology IEST-RP-DTE046.1 anti-aliasing filter See filter, anti-

18、aliasing. apparent weight/apparent mass A term sometimes used for the preferred term, effec-tive mass (see mechanical impedance). aperture error See error, aperture. Argand diagram See Nyquist diagram. array An ordered set of numbers; in a sampled data system discussed in this Glossary (see sampling

19、) the num-bers are equally spaced in time or frequency. autocorrelation function A measure of the similarity that a function has with a time-displaced version of itselfas a function of the time displacement or lag. When the lag is zero, the value of the autocorrelation is equal to the mean square va

20、lue of the function. The sample autocorrela-tion function )(jRxof the sequence xiis usually computed =+=jNijiix xxjNjR11)( j = 0,1,.,m where xi= the function x(t) sampled at the time it t = the sample interval jt = the lag N = the number of samples of x mt = the maximum lag For a discrete stationary

21、 random quantity where N must be finite, the formula gives an estimate only with a statistical uncertainty that increases as N de-creases. It is used to identify periodic components in the sample function xi. autospectral density function See power spectral density function. average, arithmetic An a

22、veraging technique that indicates the average value after N summations: x =1NxiN10 Note that xisteadily grows until i = N, at which point it is divided by N. average, ensemble A technique for averaging a set of data records. Each record (a member of the ensemble) is sampled at the same instant of ti

23、me; the sampled values are then averaged. For an ergodic random process (see ergod-ic process) an ensemble average is equivalent to the time average along any record of the ensemble. average, exponential An averaging technique that weights newer informa-tion more heavily than old. Computed as: x i+1

24、 x i+xN x iKwhere x i= old average xN= latest value K = averaging weighting factor average, stable An averaging technique which indicates the esti-mated value at each measurement as opposed to an arithmetic average which gives the values only after division by the number of averages. Computed as: x

25、 i+1= x i+xN x ii +1where i x = old average xN= latest value (N = 1 to M) i = current count index (O to M) M = total number of averages bandpass filter See filter, bandpass. bandreject filter See filter, bandreject. bandwidth, 3-db (b3db) The interval between the upper and lower frequencies at which

26、 the filter attenuates an applied signal by 3 dB (see decibel) with respect to the gain at the cen-ter frequency of the filter. bandwidth, equivalent noise See effective (noise) filter bandwidth. bandwidth, real time The range of frequencies over which a real time anal-ysis can be performed (see rea

27、l time analysis). baseband analysis Analysis over a frequency range from zero to a speci-fied upper frequency (see zoom analysis.) Bessel filter See filter, Bessel. IEST 2011 All rights reserved Institute of Environmental Sciences and Technology IEST-RD-DTE046.1 5 bias A tendency of a statistical es

28、timate to deviate in one direction away from a true value. bias error See error, bias. block See frame, memory length. block, synchronous A data sampling or processing technique in which the beginning or ending of a data block is synchronized with an external event. blocked constraints See mechanica

29、l impedance/mobility. boxcar function See window. Butterworth filter See filter, Butterworth. calibration factor (CF; also called scale factor) Data acquired by an analog-to-digital converter (ADC) is in terms of binary counts representing some fraction of a specified full-scale voltage input. The c

30、alibration factor (CF) is the multiplier that must be applied to the digitized (see digitize) value to convert it to engineering units. For example, an acceleration signal of 5 g/V is to be digitized by a 12-bit (11 bits + sign) ADC having a full input of 10 V. The calibra-tion factor is CF = 5gV10

31、V211counts= 24.414 103g /count An ADC count of 1562 would then be equivalent to g = (1562) (24. 414 10-3) = 38.135 g center frequency A characteristic of a bandpass or band reject filter or a constant percentage filter of these types, the center frequency is the geometric mean of the upper (fU) and

32、the lower (fL) cutoff frequencies: fc= fLfUFor a constant bandwidth filter, the center frequency is the arithmetic mean (see average, arithmetic) of the upper and lower frequencies: fc=12(fL+ fU)Chebyshev filter See filter, Chebyshev. chi square See goodness of fit. circular convolution (periodic co

33、nvolution) The periodic convolution ziof two time sequences xiand yj(i = 0, 1, 2, N-1) is defined as zi= x mm=0N1yimwhere x mis a periodic extension of the sample set xm, i, e., x i+kn= xi(n = 0, 1, 2, N-1) (k = 0, 1, 2 ). The inverse discrete Fourier transform (DFT-1) of the product of two DFTs is

34、a periodic convolution of the original time sequences (see DFT, wrap-around error). circular correlation A correlation function computed in terms of a circu-lar convolution. clipping The term applied to the generally undesirable (but sometimes intentional) circumstance when an output signal is limit

35、ed in some sense by the full-scale range of an amplifier, ADC, or other device. Clipping may be hard, that is, when the signal is strictly limited at some level; or it may be soft, in which case the clipped signal continues to follow the input at some reduced gain (see gain). co (coincident, real) T

36、he real part of a complex function, that is, the com-ponent that is in phase with the input excitation. In frequency domain analysis, the coincident terms are the cosine terms of the Fourier transform (see fre-quency domain, Fourier transform). coherence function A frequency domain function generall

37、y computed to show the degree of a linear, noise-free relationship between a system input and output. Values of cohe-rence satisfy the relationship O fl. on-line processing See real time analysis. ordinary coherence See coherence, ordinary. overlap processing The concept of performing a new analysis

38、 on a segment of data in which only a portion of the signals has been updated (some old data, some new data). (See FFT.) padding with zeros To append zero values on the end of an array. Used when taking an FFT of an array whose length is not a whole power of two when the algorithm requires a power-o

39、f-two array size. Also, used to avoid circular convolution errors (see FFT, circular convolution, double length transform). partial coherence See coherence, partial. passband analysis Analysis of signals (information) that occur in a known, usually restricted bandwidth, Normally applies to frequency

40、 domain analysis, which does not include zero. (See baseband analysis, frequency domain.) peak channel hold (peak hold) A frequency domain averaging method that saves the highest response measured in each quantisized incre-ment during a specified time internal or number of spectral averages. The res

41、ultant spectrum is a compo-site of the highest spectral values measured during the averaging process (see frequency domain). peak level hold (overall level) A frequency domain measurement where a series of spectra are compared as to overall level (total power in band of interest, usually RMS) and th

42、e spectrum with the highest overall level is retained (see frequency domain, RMS). peak spectra A frequency domain measurement where, in a series of spectral measurements, the one spectrum with the highest magnitude at a specified frequency is retained. periodic convolution See circular convolution.

43、 periodic in the window Term applied to a situation where the data being meas-ured in a sampled data system is exactly periodic (re-peats an integer number of times) within the frame length. This results in a leakage-free measurement in digital analysis instrumentation if a rectangular win-dow is us

44、ed. Real signals are typically not periodic in the window unless sampling is synchronized to the data periodically. IEST 2011 All rights reserved Institute of Environmental Sciences and Technology IEST-RD-DTE046.1 17 periodic random noise A type of noise generated by digital measurement sys-tems tha

45、t satisfies the conditions for a periodic signal, yet changes with time so that devices under test re-spond as though excited in a random manner. When transfer function estimates are measured with this type of noise for the excitation, each individual measure-ment is leakage-free and, by ensemble av

46、eraging, the effects of system nonlinearities are reduced, thus pro-viding benefits of both pseudo-random and true-random excitation. periodic waveform A waveform that repeats itself over some fixed period of time. periodicity The r repetitive characteristic of a signal. If the period is T (sec), th

47、en this results in a discrete frequency or line spectrum with energy only at frequencies spaced at 1/T (Hz) intervals. phase angle The time displacement between two sinusoidal quanti-ties measured relating to the time of one complete cycle of the sinusoidal. The phase can be expressed in degrees (or

48、 radians) where 360 (or 2) represents one complete cycle. A phase angle can also be defined as the angle, A, given by A = tan-1(y/x) where x and y are the real and imaginary parts of a complex number. poles The roots (possibly complex) of the denominator of a ratio of polynomials (possibly complex).

49、 Poles are used in continuous or discrete system transfer function analysis and describe the impulse response characteris-tics of systems (see zeros). power A function that varies as the square of a measured quantity. Power in this sense is not necessarily physi-cal power. power spectral density (PSD) function Also called the auto spectral density; a real-valued continuous function of frequency, presented with fre-quency on the horizontal axis and density on the ver-tical axis. A special case of a cross-