Scalar Calibration of Aeromagnetic Data Using BPANN and LS Algorithms Based on Fixed-Wing UAV Platform

Abstract

A new airborne platform named the unmanned aerial vehicle (UAV) is used to detect the geomagnetic anomaly with low magnetic interference. Nevertheless, there are steering errors of three-axis fluxgate magnetometers (TFMs) with the changes of UAV flight course. The main reasons are due to the effects of nonorthogonality, scale factors, and zero shifts. Therefore, it is quite necessary to establish calibration methods to get magnetic information of high precision. The methods of least squares (LS) and backpropagation artificial neural network (BPANN) are proposed to correct the system errors of measured data in this paper. The results show that the errors are suppressed using LS and BPANN methods. The measured errors of geomagnetic field decrease obviously after calibration. Furthermore, the BPANN method is more effective to calibrate the data error of TFMs than LS method when UAV changes its flight direction. Moreover, the spatial distributions of magnetic fields for TFMs after calibration using LS and BPANN methods are quite consistent with the distributions for optical pump magnetometers. This paper can provide a better way to improve the performance of TFMs and be widely used in the data error calibration of multiaxis sensors.