P6.2 - On Measurement and Signal Processing of Low Frequent Magnetic Fields
- Event
- SENSOR+TEST Conferences 2009
2009-05-26 - 2009-05-28
Congress Center Nürnberg - Band
- Proceedings SENSOR 2009, Volume II
- Chapter
- P6 - Sensor Electronics/Wireless
- Author(s)
- C. Rueckerl - Forschungs- und Transferzentrum Leipzig e.V., Leipzig, Germany
- Pages
- 427 - 432
- DOI
- 10.5162/sensor09/v2/p6.2
- ISBN
- 978-3-9810993-5-5
- Price
- free
Abstract
The aim of this article is to facilitate automatic treatment of three-dimensional measurement data.Especially the measurement of low frequent magnetic fields with an isotropic field probe can be seen as an example for higher dimensional sensor data.
Magnetic fields are measured and judged according to their possible negative influences on the human body. For a meaningful evaluation of non-sinusoidal fields some methods are applied in frequency domain, others on estimation of time derivative or the absolute change of magnetic flux density, which are responsible for induced body currents. These are restricted by norms and guidelines. Induced body currents can easily be calculated with measured field data of magnetic flux density and a plain loop model of human body. Therefore, a one-dimensional time series is required. The threedimensional time series of measured data has to be transformed into a one-dimensional one. If only one source is assumed, this can easily be realized with a special signal processing after or during measurement which is based on the eigenvalues of the sample covariance matrix. If there are two ore more sources superimposing the fields can be separated with automatic procedures. The use of blind source separation algotithms like Independent Component Analysis is one of the possible alternatives.
The introduced methods enable the automatic detection of superimposed fields which can be implemented in measurement equipment. This feature can help the user to valuate the field situation on the spot (e.g. in a working area). The introduced methods will be applicated to measurement data in two examples.