Abstract: In mission of space situational awareness, the inertia matrix changes greatly due to the fast fuel consumption, while spacecraft maneuver fast to observe the target in close range. Extended Kalman Filter is modified to identify the time varying inertia matrix in real time when spacecraft maneuver. Firstly, a linear state observer is designed to estimate the angular accelerations of spacecraft. Then, considering the variation of the moment of inertia of the spacecraft, the covariance matrix of prediction error is modified. A filter model is obtained with angular acceleration estimates for time varying inertia matrix identification in real time. Finally, the traditional PD like controller is used for attitude tracking with inertia matrix identification, and the attitude tracking and parameter identification are achieved.

Key words: space situational awareness, time varying moment of inertia, extended kalman filter, parameter identification, state observer