Fault detection and isolation of control moment gyros for satellite attitude control subsystem

Abstract

Abstract Control moment gyros, as one of the most commonly used actuators onboard satellites, are prone to faults and failures. The ability to detect faults, isolate their location, and identify their severity can enhance mission success rate and reduce maintenance and damage costs extensively. Therefore, in this paper, a model-based fault detection, isolation, and identification scheme is proposed and evaluated. Firstly, an adaptive threshold fault detection algorithm is proposed, using Unscented Kalman filters in conjunction with residual and innovation sequences. Secondly, a fault isolation and identification algorithm is proposed using a binary grid search method to adapt joint estimation Kalman filter’s covariance matrix once a fault is detected. Extensive Monte Carlo simulations are conducted to evaluate the performance of the proposed schemes for a 3-axis stabilized satellite with a four-single-gimbal control moment gyro cluster. Results show superior performance of the proposed methods with faster tracking and more accuracy.