State estimation for solid apogee motor burning of geostationary satellite

Abstract

This paper discusses the results of the state/orbit sequential estimation from the real doppler data of 0.1 s sampling measured for the short time boost phase of about 40 s by the solid apogee motor mounted on the Japanese geostationary satellite; broadcasting satellite, communications satellite and the geostationary meteorological satellite. The estimations were performed mainly by two stochastic Kalman filters. The first is the point mass model filter. The second is the attitude motion model filter, newly developed by the simplification of the detailed estimation algorithm, especially for the oscillating data with a spin period of the satellite equipped with an offset antenna. The detailed algorithm was generally constructed based upon the satellite spindynamics during apogee motor burn, considering the satellite mass property changes, C.G. off-set, and the motion of an offset antenna. The estimated state at motor burn out is in good agreement with the drift orbit and attitude independently determined by the ordinary Batch method. Particularly, the convergency of error covariance is excellently improved. This sort of analysis is important not only on the precise evaluation of the occurred apogee motor firing error, but also on the contribution to the real time estimation for the prompt maneuver decision just after solid motor firing, especially for the liquid apogee motor satellite.