Kalman Filter Innovation Sequence Based Fault Detection in LEO Satellite Attitude Determination and Control System

Abstract

In this paper, fault detection algorithm for LEO satellite attitude determination and control system using an approach for checking the statistical characteristics of Extended Kalman filter (EKF) innovation sequence is proposed. It is based on given statistics for the mathematical expectation of the spectral norm of the normalized innovation matrix of the EKF. The attitude dynamics of the LEO satellite as an example, is considered, and detection of various sensor faults affecting the mean and variance is examined.