Optimizing Low-Frequency Mode Stirring Performance Using Principal Component Analysis

We formulate and perform principal component analysis (PCA) of mechanically stirred fields, based on tuner sweep data collected across a frequency band at a single location of a sensor inside a reverberation chamber. Both covariance- and correlation-based PCA in undermoded and overmoded regime are p...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on electromagnetic compatibility 2014-02, Vol.56 (1), p.3-14
Main Author: Arnaut, Luk R.
Format: Article
Language:English
Subjects:
Angular position
Applied sciences
Communication, education, history, and philosophy
Computers in experimental physics
Correlation
Data acquisition: hardware and software
Direct energy conversion and energy accumulation
Electrical engineering. Electrical power engineering
Electrical power engineering
Electromagnetic compatibility
Electronics
Exact sciences and technology
Finite element analysis
Frequency measurement
Handbooks, dictionaries, tables, and data compilations
Instruments, apparatus, components and techniques common to several branches of physics and astronomy
Loading
Manganese
Miscellaneous
Multi-variate analysis
Nonuniformity
Optimization
Physics
Physics literature and publications
Principal component analysis
Principal components analysis
reverberation chamber
statistical modelling
Stirrers
Testing, measurement, noise and reliability
Tuners
Vectors
Zinc
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We formulate and perform principal component analysis (PCA) of mechanically stirred fields, based on tuner sweep data collected across a frequency band at a single location of a sensor inside a reverberation chamber. Both covariance- and correlation-based PCA in undermoded and overmoded regime are performed and intercompared. The nonstationarity of the stir performance as a function of the angular position of the stirrer is demonstrated. It is shown that this nonuniformity can be quantified and exploited to select a set of optimal stir angles. The rotated principal components are found to be interpretable as energy stirred by specific angular sectors of the stirrer and are related to the correlation structure of the data. The analysis leads to the concept of eigen-stirrings (stir modes), which form an orthonormal set of empirical basis functions for expanding stir data.
ISSN:0018-9375
1558-187X
DOI:10.1109/TEMC.2013.2271903