Complete Calibration of Three-Axis Strapdown Magnetometer in Mounting Frame

Three-axis magnetometers (TAMs) are widely used in connection with vehicle navigation systems and many other engineering applications. Technological limitations in sensor manufacturing and unwanted magnetic fields will corrupt the measurements of TAMs. In this paper, a new method is proposed to calibrate strapdown magnetometer in sensor frame. A complete error model, including instrumentation errors (body frame misalignment, scale factors, non-orthogonality, and offsets) and magnetic deviations (soft and hard iron), is elaborated. The calibration process is divided into two steps. The first step is providing a consistent solution to the ellipsoid fitting problem using a non-linear least squares estimator and genetic algorithm. The experimental system mainly consists of tri-axis HMC5843 AMR sensor and proton magnetometer. The true scalar value of magnetic field was obtained with proton magnetometer. Simulated results show that the error average of nonlinear least square is much lower than the genetic algorithm. The results of the calibration algorithm in the first step provide magnetic readings in the calibration frame. In the next step, the three calibrated axes of the given magnetometer, which are nominally orthogonal, are aligned into a true orthogonal sensor mounting frame by offering a rigorous vector analysis of rotational kinematics named vector product operations. In order to determine the magnetometer misalignment in the mounting frame, a surface plate is used as the main reference plate. Therefore, the calibration and alignment methodology is formulated in the sensor frame, without resorting to external information or models about the magnetic field.