Agile Micro-nano Satellite Attitude Estimation with Lost Measurements

Abstract

Abstract Aiming at the problem of lost measurement in attitude determination process of the agile micro-nano satellite, a novel algorithm is proposed employing the Bernoulli process to describe the loss process. Firstly,the model of gyro and star sensors as well as the satellite attitude estimation system simplified with gyro is given. Since the missing observations, the modified EKF algorithm is derived by combining a Bernoulli random variable and a real-time measurement noise correction algorithm. Then the convergence of the improved algorihtm is analyzed, and the bottom boundary condition for the observation arrival rate is presented. Finally, the improved method is compared with the conventional one through digital simulation, and the experimental results show that the preciseness of the privious one is one or two orders improved compared with the latter.