This report presents a comprehensive study on the design and control of an overactuated quadcopter in simulation and hardware. The vehicle is engineered to autonomously maneuver intricate flight paths with better performance than a traditional quadcopter. To accomplish this, we designed a novel hardware system consisting of four thrust motors and four tilt motors, corresponding to each of the drone’s four arms. This setup allows the drone to change the direction of motion more smoothly while minimizing its roll, pitch, and yaw, leading to more precise control in narrower spaces. We developed a custom controller using the Model Predictive Control (MPC) algorithm. The MPC controller generates angular velocities for the drone’s brushless motors, and tilt angles for the servo motors based on an input trajectory. The algorithm enables real-time optimization of control inputs, allowing the drone to adapt to a dynamic environment robustly. This algorithm was tested in simulation with a trajectory that mimicked a real-life slalom course on a traditional quadcopter and our drone. Keeping all other variables the same, the results of this experimental evaluation highlighted the superiority of the overactuated system in terms of maneuverability, stability, and trajectory tracking accuracy. Our research also delves into the detailed design aspects of the overactuated UAV, including the body design, selection of appropriate materials, propulsion mechanisms, and control electronics. The hardware design was tested with the onboard MPC controller, making it perform consistent hover with real-time state feedback. In conclusion, this research presents a novel overactuated UAV with the entire software and hardware pipeline designed from the ground up by us. The presented findings contribute to the advancement of UAV technologies, with potential applications in surveillance, search and rescue, and precision agriculture, among others. This work serves as a valuable reference for researchers and engineers seeking to enhance UAV maneuverability and control in autonomous flight scenarios.




Download Full History