Back-stepping calibration algorithm for three-axial magnetometer applied to autonomous underwater vehicles with magnetic deviations

Heading estimation is one of the main challenges in low-cost inertial navigation systems (INSs). Non-observability of heading angle with gravitational acceleration vector as well as inaccessibility of radio/satellite navigation in underwater vehicles increases the value of this challenge. Applying three-axis magnetometer and heading estimation from earth magnetic field components is one of the main approaches to accuracy enhancement of the low-cost inertial navigation systems. However, in order to achieve accurate heading estimation, the magnetometer must be appropriately calibrated for both sensor errors and presence of magnetic deviations. This paper aims to develop back-stepping algorithm for magnetometer calibration applied to measure the earth magnetic field components. In the proposed algorithm, the results of the prevalent spherical magnetic calibration are corrected based on vertical channel decomposition and magnetic field components in the horizontal plane. The algorithm is evaluated in the field tests executed on an Autonomous Underwater Vehicle (AUV).