Ferromagnetic Object Localization Based on Improved Triangulating and Ranging

The performance of scalar triangulation and ranging (STAR) for the localization of ferromagnetic objects is limited by asphericity error. Alternatively, iterative strategies may be used to reduce error and improve performance. However, iterative methods have problems, such as multiple required iterations, singularities, and limited accuracy. Eigenvalue methods are primarily based on the property of gradient tensor eigenvalues. A rotationally invariant scalar related to eigenvalues can eliminate the asphericity error. Additionally, a new strategy for estimating the length of the bearing vector is simpler and can be used to avoid singularities. With these two improvements, the formula for estimating the bearing vector is more concise and direct. The results show that the eigenvalue method performs better than the STAR and iterative methods.

I. INTRODUCTION

Magnetic anomalies generated by a ferromagnetic target can be adequately modeled as an equivalent magnetic dipole moment at a distance. Triaxis magnetometers (TAMs) are commonly used to locate ferromagnetic objects, which have many applications, such as unexploded ordnance detection [Salem 2005, Beran 2008, Abdelazeem 2016], biomedical applications [Nowak 2003, Poh 2010], and indoor localization [Liu 2017].