Adaptive Kalman filtering, failure detection and identification for spacecraft attitude estimation

Future space missions call for unprecedented levels of autonomy, reliability and precision, thereby increasing the demands on spacecraft failure detection, identification and compensation (FDIC) systems. We address the problems of spacecraft attitude determination (AD) and FDI for sensors and actuat...

Full description

Saved in:
Bibliographic Details
Main Authors: Mehra, R., Seereeram, S., Bayard, D., Hadaegh, F.
Format: Conference Proceeding
Language:English
Subjects:
Actuators
Adaptive filters
Bayesian methods
Covariance matrix
Extraterrestrial measurements
Fault detection
Filtering
Kalman filters
Maximum likelihood estimation
Space vehicles
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Future space missions call for unprecedented levels of autonomy, reliability and precision, thereby increasing the demands on spacecraft failure detection, identification and compensation (FDIC) systems. We address the problems of spacecraft attitude determination (AD) and FDI for sensors and actuators by developing: 1) a nonlinear extended Kalman filter (EKF) which does not require a small angle approximation; 2) a method for online tuning of noise covariances; and 3) a multi-hypothesis extended Kalman filter (MEKF) for detection and identification of sensor (gyro, Star tracker) and actuator (thruster) failures. A nonlinear EKF is designed for AD using angular rates, quaternions and the gyro biases as state variables. It is shown to provide more accurate estimates of the attitude angles and can be used for the detection and removal of bad gyro and Star-tracker measurements. The MEKF approach is used for the detection and identification of gyro, Star-tracker and thruster failures. Gyro failures are the quickest to detect and identify, followed by thruster and star tracker failures.
DOI:10.1109/CCA.1995.555664