An Eigenstructure Assignment Embedded Unknown Input Observe Approach for Actuator Fault Detection in Quadrotor Dynamics

Abstract Application of Unmanned Areal Vehicles for both civilian and military demands improved safety conditions to avoid potential malfunction and accidents in critical mission deployment. This paper presents a method for fault detection and identification (FDI) of actuator fault of a quadrotor. A combination of an Unknown Input Observer (UIO) and Beard Basic Fault Detection Filters (BFDF) are used to generate robust and directional residual using unknown input and eignestructure assignment respectively for fault identification and isolation. The uni-directional behavior of the residual will be exploited to isolate the source of the fault by comparing with known or predefined fault directions. The actuator faults are modeled as a loss of effectiveness, Lock-In-Place, Float and Hard Over Failure. The FDI system is used to detect and isolate the actuator faults in quadrotor actuator ( motors). A numerical simulation is done to demonstrate the effectiveness of the UIO and BFDF based FDI algorithm on a model quadrotor.