Abstract

ABSTRACT
Development of Fault Detection and Diagnosis Techniques with Applications to Fixed-wing and Rotary-wing UAVs
Ling Ma
Fault Detection and Diagnosis (FDD), as the central part of a Fault Tolerant Control System (FTCS), detects and diagnoses the source and the magnitude of a fault when a fault/failure occurs either in an actuator, sensor or in the system itself. This thesis work develops an applicable procedure for a FDD scheme to both fixed-wing and rotary-wing UAVs (Unmanned Aerial Vehicles) in the discrete-time stochastic domain based on the Kalman filter techniques. In particular, the proposed techniques are developed in highly nonlinear and 6 degree-of-freedom equations of Matlab/Simulink simulation environment for a quad-rotor helicopter UAV, a Boeing 747, and a NASA Generic Transport Model (GTM) fixed-wing UAV. A key development in this thesis is that an Adaptive Two-Stage Extended Kalman Filter (ATSEKF) algorithm and a Dual Unscented Kalman Filter (DUKF) algorithm are applied for simultaneous states and fault parameters estimation of these UAVs. The statistical decision-making techniques for fault detection and diagnosis are also discussed in the presence of partial faults in the UAVs. The measured system outputs and control signals are used as inputs of the ATSEKF and DUKF, and the estimated states and parameters are used for comparison and analysis in the fault detection and diagnosis. The simulation results show that the effectiveness and performance of ATSEKF and DUKF for the purpose of fault detection and diagnosis of both fixed- and rotary-wing UAVs are satisfactory.