Relative navigation for a malfunctioned spacecraft by using an adaptive state observer

This paper proposes an adaptive state observer to estimate the relative position and velocity between a malfunctioned spacecraft and a servicer by using a stereovision system. Firstly, a new visual measurement algorithm is presented according to the geometrical relations of a stereovision system, and the CW equations are utilized as the state equations. Secondly, a full-state observer is designed to estimate the relative position and velocity instead of a Kalman filter, avoiding modeling for stereo vision measurement noise and process noise. Thirdly, because the convergence speed and estimating accuracy are related to negative real parts of pole, an adaptive state observer is proposed to improve the rate of convergence and estimating accuracy. Because adjusting time is certain for a linear time-invariant system, the adaptive state observer is realized by utilizing a time switch. Simulation results demonstrate the effectiveness of the algorithm. The full-state observer with an adaptive pole switch can effectively improve the accuracy and speed up the convergence speed of relative navigation.