Model Predictive Controller (MPC) Design for an UAV Quadcopter in Windy Critical Environment

• Main Author: Mustapha Amine, Sadi
• Format: Thesis
• Language: English
• Published: e-Prime - Advances in Electrical Engineering, Electronics and Energy 2025
• Subjects: TJ Mechanical engineering and machinery
• Online Access: http://ir.unimas.my/id/eprint/47742/; https://www.sciencedirect.com/science/article/pii/S2772671124004169

Unmanned Aerial Vehicles (UAVs), particularly quadcopters, have become indispensable in navigating critical windy environments, offering unparalleled access to complex terrains. These challenging environments pose unique navigational difficulties due to fluctuating wind patterns, which significantly impact UAV performance and accuracy. This research enhances UAV navigation effectiveness and stability through the development of a robust controller—specifically, a cascade Model Predictive Control (MPC) algorithm, optimized for continuous gust conditions as modelled by the Dryden wind turbulence approach. To accurately capture UAV dynamics within these environments, a mathematical model was developed using the Newton-Euler technique. This approach enabled precise analysis of both translational and rotational motions, factoring in elements such as inertia, torque, and external forces like drag and lift. The cascade MPC was further validated through real-time experiments, adapting UAV dynamics to unpredictable turbulence patterns and obstructions. This system’s adaptability and precision are contrasted with a Proportional-Integral-Derivative (PID) controller, highlighting MPC's predictive advantages for dynamic environments. Experiments were conducted in mangrove forests, settings characterized by complex wind patterns, using a UAV equipped with a Pixhawk hardware controller. The UAV followed predefined paths to collect data on altitude, roll, pitch, yaw, and responses to wind turbulence, enabling refined control strategies and model improvements based on real-world performance. Comparative analysis shows that the quadcopter, when equipped with MPC, not only achieved better stability and responsiveness but also improved operational efficiency under adverse weather conditions, demonstrating the system’s robustness and applicability to challenging UAV operations. This research thus advances UAV deployment in demanding real-world applications by integrating sophisticated dynamic modelling with innovative control techniques.