Motion estimation of a rigid body with an EKF using magneto-inertial measurements

Abstract

This paper studies the attitude and position estimation problem of a rigid body equipped with inertial and magnetic sensors. First a continuous-time model is provided linking the motion of the body and sensor measurements, which depend on the total magnetic field and measurement noise. The magnetic field may depend both on the space variable (since the indoor navigation is considered) and on the time variable (since power-line interference is taken into account). An Extended Kalman Filter is designed aiming at the minimization of the drifting error on the estimated position. A numerical simulation illustrates the motion estimation result. Some indoor tests are finally performed to confirm the approach and the obtained convergence on real experiments.