Intelligent Adaptive Control of a Quadcopter

• Grant number: 23/05069-0
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: September 01, 2023
• End date: December 31, 2024
• Field of knowledge: Engineering - Electrical Engineering - Industrial Electronics, Electronic Systems and Controls
• Principal Investigator: Roberto Santos Inoue
• Grantee: Gabriel Andreazi Bertho
• Host Institution: Centro de Ciências Exatas e de Tecnologia (CCET). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil
• Associated scholarship(s): 23/18487-5 - Experimental Validation of Deep Learning-Based Adaptive Control for a quadcopter, BE.EP.IC

Abstract

Unmanned Aerial Vehicles (UAVs), also known as Drones, often operate under adverse environmental conditions, where external disturbances, such as wind gusts, and parametric uncertainties in the dynamic model of these vehicles - as mass, center of gravity, and moments of inertia - are constantly present. In this scenario, it is important that the control used is robust enough to deal with such disturbances, ensuring safety, stability, and efficiency of the aircraft. In this Scientific Initiation (SI) work, we chose to develop an intelligent model reference adaptive control to ensure the robustness of the system of a quadcopter drone considering the aforementioned scenario, as well as to track its trajectory. To this end, deep neural networks will be used to learn the parametric uncertainties and estimate the external disturbances of this aircraft.The implementation of the control architecture will be developed through Python and the TensorFlow framework. For experimental validation, we intend to implement the simulated architecture using the driver bebop_autonomy for the framework Robot Operating System (ROS) framework on a commercial quadcopter, the Parrot Bebop 2, which already has stabilization control. Finally, the final validation of the system will be done using the state-of-the-art drone simulation tool Parrot Sphinx.