Abstract:
The control and navigation of Unmanned Aerial Vehicles (UAVs) is quite demanding in general as these autonomous vehicles experiences different kind of faults during flight. Successful flight of UAVs hence demands some methods or techniques by virtue of which these autonomous vehicles can cater for these faults and deal with them efficiently as early as possible. Many different schemes have been developed for the Fault Tolerant Control (FTC) of UAVs. In this study, a robust approach for the FTC of a UAV in the occurrence of three different forms of actuator faults namely abrupt, incipient and intermittent is developed and verified. The Fault Detection and Isolation (FDI) is achieved by employing the Observer based residual scheme. In order to make the generated residuals insensitive to disturbances and uncertainties, the method of Eigenstructure Assignment (EA) is also incorporated in the FDI scheme. The Sequential Probability Ratio Test (SPRT) is employed for the efficient statistical testing of the residuals during the detection phase. Afterwards the Multiple Models Switching and Tuning (MMST) technique based on Linear Quadratic Gaussian Regulator (LQG) is designed and used as the reconfigurable controller. A linearized lateral directional model of the Aerosonde UAV is then used to test the robustness and efficacy of the proposed scheme for FTC in Simulink (MATLAB). The results of the simulations duly justify the efficiency of the developed FTC scheme for a UAV. The proposed FTC scheme is hence an efficient, robust and practical methodology for the detection and reconfiguration of the actuator faults occurring in UAVs during flight.
