Unscented transform-based uncertainty analysis of rotating coil transducers for field mapping

Area of Science:

• Metrology and Measurement Science
• Electromagnetism and Magnetic Field Measurement

Background:

• Accurate magnetic field mapping is crucial for various scientific and engineering applications.
• Traditional uncertainty analysis methods (e.g., Guide to the Expression of Uncertainty in Measurements - GUM) can be limited by approximations for nonlinear models.

Purpose of the Study:

• To develop and validate a robust methodology for analyzing the uncertainty of rotating coil transducers used in magnetic field mapping.
• To compare the proposed method with existing standards, particularly for nonlinear measurement scenarios.

Main Methods:

• Integration of the unscented transform (UT) with statistical design of experiments (DoE).
• Application of UT for propagating uncertainties in nonlinear measurement models, avoiding linearization approximations.
• Utilizing DoE for efficient exploration of uncertainty sources and their contributions.

Main Results:

• The UT-based approach provides a more error-proof analysis for nonlinear magnetic field measurement models compared to GUM linearization.
• The combined UT and DoE method significantly reduces computational burden versus Monte Carlo simulations.
• Experimental validation showed that reducing coil area uncertainty by one order of magnitude decreased transducer uncertainty by a factor of 25.

Conclusions:

• The proposed unscented transform and design of experiments framework offers a superior, efficient, and accurate method for magnetic field transducer uncertainty analysis.
• This approach is particularly advantageous when GUM assumptions are not met, providing reliable results with reduced computational cost.