A Master of Science Thesis in Mechatronics Submitted to the School of Engineering by Mahmoud Hadi, "Low Cost Autopilot Design Using Fuzzy Supervisory Control," May 2005. Chair of Committee Dr. Mohamed Ameen Al Jarrah & Co-Advisor Dr. Abdel Rahman Al-Ali. Available are Soft and Hard Copies of the Thesis.
With the recent development in single-board embedded systems, wireless communications technologies and mechatronics systems, aerospace researchers have an opportunity to design an affordable controller that can enhance the performance of Unmanned Aerial Vehicles (UAVs). The controller has to be able to drive the nonlinear aerodynamic system with maximum stability and fast response. In this research, different control methods have been studied and it is decided to use a nonlinear control method to help adapting the high nonlinearity in our system. The high nonlinearity is due to the coupling between the nonlinear dynamic aircraft model and the nonlinear aerodynamics inputs. Therefore a nonlinear flight controller using Fuzzy Gain Scheduling is implemented and simulated on Matlab Simulink. Fuzzy system is used to interpolate between different linear control laws to maintain aircraft stability. The simulation of the LQR FGS (Linear Quadratic Regulator, Fuzzy Gain Scheduling) shows that the aircraft preformed a stable flight compared to the LQR controller. The later was not able to keep a stable flying condition for the aircraft for the same input command and initial conditions; this shows the advantage of using nonlinear controllers over linear controllers for such systems. In this thesis, a fuzzy interface was designed to make this nonlinear control algorithm suitable for the proposed avionics unit. A low cost avionics unit was also manufactured using off-the-shelf stand-alone embedded system board based on 16-bit Microcontroller. The main concern in developing this avionics unit is to design, develop, implement and test an affordable, reliable, and programmable Digital Flight Control System (DFCS). The DFCS contains a variety of Off-The-Shelf commercial sensors that enables it to get all the states needed to fully control the aircraft. The price of the developed avionics unit was less than $1000, which is very low compared to the available avionics units.