Enhancing Quad-rotor Maneuverability with Adaptive Intelligent Control

In recent years, the use of unmanned aerial vehicles or drones in a multitude of fields has significantly increased. The most utilized type of drone can be considered as the quad-rotor, which is a drone with 4 active rotors used. The scope of study for drones can range from the field of agriculture to indoor scanning and construction, and making them an appealing research topic. The current research primarily focuses on validating the usage of drones in a certain field, and testing different control methods as to try and find the most suitable approach for the most generic of usages. The goal is to address an aspect of testing regarding different drone control methods which has yet to be thoroughly examined. With further contemplating, the missing research can be identified as a lack of different environmental conditions. To validate the identified controller, tests are carried out both linear and highly non-linear environments. This is done to acquire a deterministic ruling on the optimum method to move forward with UAV control methods. PID control and FLC were the control methods opted to be applied and both are tested in 2 Gazebo/ROS simulation environments, with the latter simulation including noise in the form of wind. The drone used is the PX4-Autopilot Iris drone. Testing has showed that FLC is better used in highly non-linear situations. This is represented in the results which highlights the
error of the FLC controller in linear situation which can reach a maximum of 3 meter error between the desired and actual final position. This is while having the PID control which has a maximum of over 20 meter difference between the desired and actual final positions with respect to the same error in the FLC controller. However, a conclusion of combining both control methods will be the optimum solution going forward.