Autonomous quadrocopter control

Quadrocopters, or quad rotor helicopters, are typically small and unmanned. New uses are still being developed for these aircraft such as autonomous delivery services or surveillance. The quadrocopter in this thesis is optimized for indoor flight. So, a safer chassis is designed to keep the propellers protected. Also, a linkage for cargo manipulation is designed and optimized using a genetic algorithm. PID controllers are designed, which rely on many different sensors to determine the present position and orientation of the craft such as accelerometers, cameras, and wireless trilateration. A simple use-case is simulated where a quadrocopter can sort blocks based on a learned color pattern using an artificial neural network (ANN). A data storage and search system is developed to quickly search for obstacles to avoid. Finally, a prototype quadrocopter is constructed to implement some of the designed components and features, such as the controller and some sensors. Many of the systems developed, like the linkage solver, wireless trilateration, and spatial data storage system, have significant uses beyond those for a quadrocopter.