Describing a variable whose value at a particular future instant cannot be predicted exactly.
A random input is uncorrelated for a lag
if for every lag greater than
the autocorrelation function is zero. White random noise (equal energy at
all frequencies, a constant PSD) will be uncorrelated for all lags greater
than zero. (See AUTOCORRELATION FUNCTION,
WHITE NOISE, PSD.)
Vibration whose instantaneous value cannot be predicted with complete certainty for any given instant of time. Rather, the instantaneous values are specified only by probability distribution functions which give the probable fraction of the total time that the instantaneous values lie within a specified range.
Notes: "random" means not deterministic, "white" means uncorrelated (flat PSD), "Gaussian" describes the shape of the probability density function, and "noise" usually means not the signal, but can mean the random process. These are all different, though related.
See FUNCTION.
In a complex signal, the component that is in phase with the excitation. In frequency domain analysis, it is the magnitude of the COSINE terms of the Fourier series, "Coincident, Co," as in COQUAD analyzer (See FREQUENCY DOMAIN).
A function whose range is the subset of real numbers (See FUNCTION).
See FILTER, DIGITAL.
Analysis for which, on the average, the computing associated with each sampled record can be completed in a time less than, or equal to, the record length. In digital analyzers, the functions accomplished during the computing time should be specified; e.g., Fourier transform, calibration, normalizing by the effective filter bandwidth, averaging, display, etc. (See FOURIER TRANSFORM).
See FILTER, DIGITAL.
The desired PSD or SRS of a control signal.
A coefficient in the numerator of the partial fraction expansion of a transfer function. The residue can be associated with the mode shape.
The discernible difference between one value and adjacent values in a measurement.
In a digital signal analyzer it usually refers to the smallest time or frequency
increment that can be discerned. In frequency domain measurements the frequency
resolution is also called delta
(
) and is equal to the analysis
bandwidth divided by the number of spectral lines measured. Since only periodic
signals can be resolved to within
,
the "effective noise bandwidth" of a digital analyzer is probably a more meaningful
measure of resolving ability in the frequency domain (See FREQUENCY
DOMAIN).
The enhancement of the response of a physical system to an excitation. The resonant frequencies are usually defined as those frequencies where a small change in the frequency of excitation in either direction will cause the system response to decrease. The term resonant frequency is also used (sometimes erroneously) to denote the imaginary coordinate of the poles of a transfer function, the undamped natural frequencies of a system, and the damped natural frequencies.
See FILTER, RINGING
The square root of the average of the square of a function taken through
one interval
:

Usually refers to a filter characteristic. The best straight-line fit to the slope of the "filter transmissibility characteristic" in the "transition band," usually expressed in dB per octave.
See ERROR ROUNDOFF.