UAS Model Identification and Simulation to Support In-Flight Testing of Discrete Adaptive Fault-Tolerant Control Laws

In mission-critical applications of unmanned and autonomous aerial systems(UAS), it is of significant importance to develop robust strategies for fault-tolerant systems that can countermeasure system degradation and consequently support the integration into the National Airspace (NAS). This thesis research illustrates the results of systems identification that is performed using DATCOM followed by the flight test data. This data is acquired from conducting an intensive flight testings program of a fixed-wing UAS to determine the state-space model of the aircraft. A discrete state-space system is reconstructed from these models to derive Auto-Regressive Moving-Average (ARMA) models used to design a Discrete Direct and Indirect Model Reference Adaptive Control. Description of the UAS, sub-systems, and integration is presented in this thesis along with analysis of results from numerical simulation to support the design, development, and validation of adaptive control laws for fault tolerance. A set of performance metrics are defined to perform the analysis in terms of control effort, tracking performance, and reconfiguration of control laws under commonly occurring failures such as partial control surface damage, pilot-induced oscillations, and uncertain ice accretion