Position Extraction of Ultralow-Speed Gimbal Servo System With Linear Hall Sensors

Abstract

This article explores a novel composite signal extraction method to extract high-precision angular position from the signals detected by Hall sensors in the gimbal servo system of single-gimbal magnetically suspended control moment gyro (SGMSCMG). Due to magnetic field distortion and other factors, the detected signals involve lots of harmonics and unequal amplitude, which decrease the position resolution accuracy significantly. In the proposed method, normalization is first performed to obtain signals with equal amplitude. Then, on the basis of offline analysis, the signal reconstructed model is established to suppress the most harmonics. Finally, the errors of the models and unmodeled noise are restrained by applying the discrete extended state observer (ESO). Simulated results show that the influence of harmonics and unequal amplitude on the angular position can be almost attenuated by the proposed method. Besides, plentiful experiments on an SGMSCMG demonstrate the validity of the proposed method, and the desired angular position of the ultralow speed servo system can be extracted effectively.