MAG-MFDN: geomagnetic vector measurement interference compensation method for autonomous vehicles

Geomagnetic navigation uses the Earth’s magnetic field for positioning and is widely used in autonomous vehicles. However, affected by the magnetic measurement interference from ferromagnetic structures and electronic devices, the navigation accuracy fails to meet the requirements for the operation of autonomous vehicles. Therefore, interference compensation for the magnetic measurement data is needed. Existing compensation methods exhibit poor performance under the influence of non-stationary electromagnetic interference (EMI) during the operation of the autonomous vehicles. To address these issues, this paper designed a compensation model for geomagnetic three-axis measurement system and a method that utilizes magnetic multi-feature denoising network to accurately remove the non-stationary EMI, while simultaneously employing robust trust region to perform robust estimation of compensation parameters. Experimental results demonstrate that compared with the commonly used algorithm total least square, it reduces the three-axis errors by 51.15%, 47.01%, and 38.79%, respectively.