Intelligent control for a fixed-wing unmanned aerial vehicle
The purpose of this thesis was to create a low cost, open source, fixed-wing unmanned air vehicle (UAV). In this thesis the development, simulation and implementation of the control algorithm will be discussed. The control algorithm will maintain the UAV stable while at the same time controlling the heading to reach the user defined waypoints. The control algorithm must be applicable to similar UAVs with minimal changes to the algorithm.
It is well established that an airplane is a nonlinear, unstable, time-varying system, and that its mathematical model is a high order differential equation. By definition we can see that the conventional PI, PID, pole placement controllers will fail in the attempt of controlling an UAV. The airplane is a perfect candidate for alternative control methods such as the adaptive control, neural network control, fuzzy control, and variations or combinations of them.
Since the control algorithm must apply to different UAVs with similar characteristics, it is assumed that the exact dynamics of the UAV are not provided. Therefore, the control must be able to adapt to different UAVs or the user should be able to easily modify the control algorithm to fit the new UAV model. Based on these requirements, the proposed control algorithms are: fuzzy logic control, model reference adaptive control and radial basis function neural network control.