Design and Control of an Autonomous Unmanned Aircraft System

Abstract

V____ Abstract Substantial progress has been made recently towards the design, control and deployment of Autonomous unmanned aircraft systems (AUAS). The objective of this thesis seeks to complement this progress by providing a platform to the autonomous and intelligent flight control systems of the unmanned aircrafts, henceforth they can be integrated into national airspaces for more civil applications. The thesis hypothesizes a full autonomous flight control system that includes takeoff, en-route navigation and landing based on a new hybrid control technique. The hybrid control is a combination of different control techniques, conventional closed-loop control, PID control, model predictive control (MPC), adaptive control and Neural Network. Initially the longitudinal controller was proposed. It included three autopilots: the speed control autopilot, the pitch orientation autopilot and altitude autopilot using (PID, l\/[PC and Neural Network) techniques respectively. The hybrid intelligent longitudinal controller was simulated and tested successfully. The PlD controller has shown a significant response in compensating the errors that are related to the control surface actuators (rise time 0.ls, settling time l.8s overshot 5.5%), while the neural controller was a very robust in the control of the overall response of the system. Then the lateral was proposed based on the criteria of the hybrid control that include integration of the classical control system with neural network controller and a new guidance system was designed. Also an auto-landing controller using fuzzy controller and Neuro-adaptive controller was proposed to provide the vertical and horizontal guidance to the aircrafl for landing approach phase. Many flight tests were carried for different flight conditions and the result were significant for the control of AUAS.