Placement and Denoising of Total Magnetic Field Sensors Onboard an AUV in Support of Geophysical Navigation

A study of the proper placement of a total field QuSpin magnetometer on a REMUS 100 AUV and required denoising for sufficient isolation of the sensor from magnetic self-noise of the AUV is presented. This study is in support of a long-term goal to aid geophysical navigation based on sensing of the total magnetic field and bathymetry. Field tests were conducted in an isolated marine environment with the magnetometer placed on a pole at various distances ahead of the nose of the AUV. First, ambient magnetic field was measured over periods of several hours at various times of the year, to determine the range of variability of the field under natural conditions. Then the impacts of the internal and external sources of magnetic self-noise on the measured field were examined. Low-pass wavelet filtering was used to digitally reduce the fluctuations caused by sources of internal and external magnetic interferences, including energized onboard electronics and the AUV’s DC motor used for its propulsion, as well as skin depth and coastline effects. Vehicle motion-induced interferences, such as induced magnetic field and eddy currents, were also examined, to determine whether a Tolles-Lawson model-based motion compensation would be necessary. A two-stage measurement and analysis procedure was developed. Placing the total field magnetometer at a minimum distance of 0.4m ahead of the vehicle’s nose appears to provide sufficient isolation if wavelet lowpass filtering is applied with a cutoff frequency of 0.1 mHz. However, significant improvement is obtained if the magnetometer is placed 0.9 m ahead of the vehicle’s nose.