Autonomous Navigation Control of a Quadcopter Based on MPC

Quadcopters excel in accessing environments inaccessible to humans, owing to their small size and light weight. This versatility has made them pivotal in search and rescue missions, terrain mapping, aerial photography, and various other applications. However, the inherent nonlinear, underactuated, and strongly coupled characteristics of quadcopters pose significant challenges for designing effective flight control systems. Consequently, understanding the dynamics and achieving autonomous navigation control have become critical research areas. This paper addresses these challenges by employing a simplified kinematic model to approximate quadcopters dynamics. By delving into its flight principles and control algorithms, the study adopts an adaptive MPC algorithm to predict the UAV's trajectory. The efficacy of this approach is evaluated through experimental validation using the Parrot Bebop 2 as a test platform. Results demonstrate that The Adaptive MPC algorithm effectively tracks the quadcopters trajectory, ensuring stable autonomous navigation. Furthermore, the algorithm's short execution time meets real-time requirements in most operational scenarios.