Nonlinear aircraft attitude and heading reference system failure detection and identification

Abstract

© 2016, American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved. Using the kinematic model of the aircraft for sensor Fault Detection and Identification (FDI) can reduce the influence of model uncertainties. Many papers have used this method to detect the faults in the aircraft Air Data Sensors (ADSs). However, the Attitude Heading and Reference System (AHRS) is assumed to be fault-free in previous studies. In this paper, both the ADS and AHRS faults are considered. The kinematic model including the ADS and AHRS faults is given. An Adaptive Three-Step Unscented Kalman Filter (ATS-UKF) is designed to deal with the sensor FDI problem. The FDI performance of the ATS-UKF is validated using both simulated aircraft data where no model uncertainties are included and real flight test data where model uncertainties and varying winds are present. Both the validations use different fault scenarios, which contains multiple and simultaneous faults. The results demonstrate that the ATS-UKF is able to detect, isolate and estimate the faults in the ADSs and AHRS.