A numerical methodology for a 6 DOF pose estimation with 3D magnetic field sensors

Abstract

Linear and angular position sensing using magnetic field sensors play a decisive role in automotive, industrial and consumer applications. Recently powerful 3D Hall sensors have become available, which enable simple measurement setups for 3D position sensing. In this paper a numerical concept for the determination of linear and angular position in the three-dimensional space is presented, which bases on an analytic model of an encoder magnet. To prevent numerical instabilities, a modified Newton-Raphson algorithm with adaptive dynamic step size control is used. Simulation results validate the applicability of our approach.