This work presents a path tracking methodology developed for a four rotor mini­helicopter, also called X4-flyer or quadrotor. Although previous work has successfully, to some degree, achieved trajectory tracking with similar UAVs (Unmanned Aerial Ve­hicles) moving at relatively low speeds and over short distances, the proposed method­ology described in this thesis works for vehicles moving at low and high speeds. The methodology combines linear and fuzzy control that enables the quadrotor to fly in the presence of disturbances and can be used with any pose (attitude) controller previously developed by other researchers. In this thesis, it is assumed that the path that the vehicle will follow is given by a set of waypoints provided by the user.

The proposed methodology comprises an attitude, a position, and a straight-line path controller. These controllers combine classic and intelligent control schemes which enable the vehicle to move under nominal ( without disturbances) and unexpected sce­narios (with wind gusts). Classic control techniques are used for cases when the vehicle is hovering or close to its intended hovering position (e.g., close to the target position) while fuzzy logic is employed for situations where the difference between the vehicle's current position and its goal is greater than a pre-specified amount. Thus, fuzzy logic is used during the vehicle's displacement (travel) while classic control is mainly used for hovering. So, the switching between these two control schemes is dynamic, based on the state of the vehicle.

The proposed methodology has been designed for path tracking in the horizontal plane (xy) and simulated in 3D. Complex curved paths are broken into straight-line segments that represent the desired path, to certain degree, based on user require­ments, the task and the desired precision of the vehicle's motion. Straight-line and complex path simulation studies are used as proof of concept to show how the method­ology reacts to diverse scenarios and different disturbances. Simulation results are used to show the performance of the proposed methodology compared with previously developed control mechanisms.