Orbital Calibration of Nanosatellite Magnetometers via Linear Two-Stage Kalman Filter

The scale factor and bias of the magnetometer must be determined in order to appropriately estimate satellite dynamics and control orientation. Temperature variation and thermal deformation of the structure cause the magnetometer scale factor and bias to vary, resulting in a rather substantial magnitude error that must be reduced in order to achieve mission requirements. This paper suggests a two-step method for orbital magnetometer biases and scale factors determination using a linear Kalman filter (LKF). The bias error and scale factor error are combined to produce equivalent no-steady bias noise, which is then deleted from magnetometer results as bias error. Each iteration of the calibration procedure has two stages: In the first stage of calibration, LKF estimates the equivalent magnetometer biases as well as three state parameters. In contrast to prior approaches, scale factors are not regarded as part of the state vector; at the second stage of the calibration procedure, three magnetometer biases and scale factors are computed using a newly proposed Kalman filter extension.