Research on Calibration Methods of Three-Axis Magnetometer Based on Vector Norm Estimation

In view of calibration methods of three-axis magnetometer, most calibration methodsuse the normalized magnetic field vector to calibrate three-axis magnetometer. In order to improve the generality and effectiveness of calibration methods, a two-step calibration method is proposed. In this work, the error model of three-axis magnetometer uses the magnetic field vector norm as a parameter. First step is used to get the initial parameters, then second step is used to solve nonlinear least squares problem with the initial parameters. Simulation results show that two-step calibration method can improve the convergence of Levenberg-Marquardt method. This method does not depend on the observed vector, and is also applicable to other three-axis sensors with the same error model. And experiment results show that this method can correctly and effectively compensate the misalignment errors, null shift errors and sensitivity errors of the three-axis magnetometer. And the magnetic field vector norm is relatively precise.