Optimal estimation of Gaofen-3B satellite attitude deviation based on echo frequency domain features

This paper proposed and verified an optimal estimation method of satellite attitude deviation in orbit based on the frequency domain features of microwave remote sensing data from the Gaofen-3B (GF-3B) satellite. The Synthetic Aperture Radar (SAR) imaging results of the Amazon forest's strip pattern were obtained at different imaging incident angles of the GF-3B satellite. Then, the inversions of the Doppler center deviation of the SAR signal obtained the corresponding beam pointing errors at different imaging incident angles. Next, a three-degree-of-freedom nonlinear least square fitting model was established between the Doppler center deviation of the SAR signal and satellite attitude deviation. Combined with the measured SAR signals, satellite attitude deviations in different dimensions (pitch, yaw, and roll) were separated and estimated. On this basis, the GF-3B satellite proceeded with an on-orbit attitude correction test. The maximum Doppler center deviation at all imaging incident angles decreased from 400Hz to 46Hz, and the residual beam pointing errors were less than 0.01°. The proposed optimal estimation method can guarantee the on-orbit attitude deviation suppression and image quality improvement of the SAR satellites.