Innovative Aeromagnetic Compensation Techniques for In-Cabin Magnetometers Amidst Complex OBE Interference

Recent advances in drone-based aeromagnetic survey technology face limitations due to existing interference models, and calibration methods cannot adequately compensate for the complex additional magnetic interference on in-cabin magnetometers generated by the on-board electronic (OBE) systems, which require the magnetometer to be mounted on a stinger or wing far from the cabin, restricting drone design and size. To solve this, we propose a compensation method for in-cabin magnetometers. Specifically, in the preprocessing stage, we propose a recursive empirical wavelet transform reconstruction (REWT-R) algorithm for more precise current denoising. Then, an iterative partial correlation screening (IPCS) algorithm is introduced to screen current data for model calibration, avoiding multicollinearity issues. Finally, we eliminate the inaccurate bandpass filtering operation in the calibration stage and improve the loss function with a decomposed weighting scheme, enhancing its global optimization and generalization capabilities. Experiments on public datasets validated the effectiveness of the proposed method at each stage. Meanwhile, compared to the latest methods, our proposed in-cabin magnetic interference compensation method reduces the root-mean-squared error (RMSE) by 15 times.