Accurate magnetic dipole moment estimation for magnetic sources in medical and robotic applications

Document Type

Article

Publication Date

4-21-2026

Abstract

This paper presents a two-stage technique for simultaneously estimating the magnetic dipole moment (MDM) and calibrating the effective point (EP) of a magnetometer. This precise characterization is crucial for many applications, including medical robotics, where accurate magnetic field sensing is essential. We directly incorporate EP calibration (EPC) into the MDM estimation process, which mitigates the errors that become prominent at close distances. Through a theoretical analysis, we demonstrate the pronounced influence of signal-to-noise ratio (SNR) and target-to-sensor distance (TSD) on the accuracy of MDM estimation, and that the misalignment between sensor and magnetic source has a comparatively lower impact. The proposed two-stage technique consists of: (1) an Inverse Problem Approach using a non-linear least-squares (NLS) optimization algorithm that refines both MDM and EP estimates from multiple measurements; and (2) a refined Direct Method that leverages the optimized EP to calculate the MDM from a single measurement. Numerical validation tests confirm the robustness of the Inverse Problem Approach to low SNR conditions and its effectiveness in mitigating source misalignment errors, achieving MDM estimation errors below 1%. Experiments conducted in a scaled laboratory environment further confirms the accuracy of the Inverse Problem Approach with a 1.6% error for N35 NdFeB magnets. Additionally, the refined Direct Method consistently provides a stable MDM estimate across varying TSDs, which is a stark contrast to the unstable results obtained when using a fixed geometric EP. This proposed approach offers a significant advancement in the accurate characterization of magnetic sources for a variety of applications in robotics and precision industrial automation.

Keywords

Effective Point (EP) Calibration, Industrial Automation, Magnetic Dipole Moment (MDM) Estimation, Medical Robotics, Precision Localization, Sensor Calibration, Single Magnetometer

Publication Title

Measurement Journal of the International Measurement Confederation

ISSN

0263-2241

DOI

10.1016/j.measurement.2026.120885

Funders

Monash University Malaysia

Volume

270

Publisher

Elsevier B.V.

Share

COinS