Overview of Methods for Analysis of Non-Stationary
Signals
A number of traditional analysis methods can be used for the analysis of nonstationary signals and they can roughly be categorized as follows:
1) Divide into Quasi-stationary segments by proper selection of analysis
window
a) Record the signal in a time buffer (or on disk) and analyze afterwards: Scan Analysis
b) Analyze on-line and store the spectra for later presentation and
post-processing: Multifunction/multibuffer measurements
2) Analyze individual events in a cycle of a signal and average over several
cycles: Gated measurements and analysis
3) Sample the signal according to its frequency/RPM variations (adaptive
data resampling): Order tracking measurements
4) Analysis using Autoregressive Signal Modelling: Maximum Entropy
No comments:
Post a Comment