- The paper presents a dual-rotor tail-sitter design that combines low-cost, accessible hardware with comprehensive aerodynamic modeling to enable robust flight performance.
- The paper employs a cascaded control architecture with real-time state estimation, achieving precise trajectory tracking and stable hovering in dynamic conditions.
- The paper provides a complete open-source package, including design files, control software, and simulation tools, to facilitate aerial robotics research and education.
An Overview of the Phoenix Drone: An Open-Source Tail-Sitter Platform for Research and Education
The paper presented focuses on the design, development, and performance evaluation of the Phoenix drone, introducing a dual-rotor micro aerial vehicle (MAV) with a tail-sitter configuration aimed at supporting research and educational activities in aerial robotics. The Phoenix drone is presented as a fully open-source platform, making available its design documents, control software, and simulation tools to the wider research community and educational institutions.
Design and System Architecture
The Phoenix drone features a dual-rotor tail-sitter design, emphasizing an integration of simplicity, efficiency, and performance. The vehicle is constructed using accessible materials and tools, including off-the-shelf components and 3D-printed parts. The structural design is focused on minimizing costs while ensuring robustness—utilizing a polyurethane foam core and ensuring easy assembly and modification. The drone is equipped with the PixRacer flight computer, which runs on PX4 middleware, a choice reflecting the intent to leverage widely-used platforms to facilitate research adaptation.
Dynamic characterization of the Phoenix vehicle is meticulous. The paper deploys a detailed aerodynamic model to capture flight dynamics accurately, addressing nuances such as propeller thrust and torque, and aerodynamic lift. The integration of this model into the vehicle's control architecture allows the Phoenix drone to achieve precise hovering and dynamic maneuver capabilities.
Control Methodology
A significant portion of the paper is dedicated to the control strategy employed by the Phoenix drone. Utilizing a cascaded control architecture with separate position, attitude, and rate controllers enables highly precise trajectory tracking and stabilization. This control architecture is supported by a real-time state estimation process facilitated through a complementary filter, integrating IMU data with motion capture inputs.
The control system is further augmented by a feedback linearization component within the control loop, ensuring seamless mapping of desired torques and forces to actuator commands. Through test flights, the architecture demonstrates robustness in maintaining stability even under varying dynamic conditions.
Performance Evaluation
Empirical results showcase the Phoenix drone's capabilities in precision hovering and trajectory tracking. The vehicle demonstrates minimal positional errors in stationary flight tests and retains its stability during complex maneuver patterns. Tests conducted in a controlled motion capture environment quantified the drone's performance, illustrating impressive agility and reliability—rendering it suitable for both indoor and outdoor applications.
The numerical results presented include RMS positional errors during hover trials, highlighting the effectiveness of the control strategy in maintaining the vehicle's stability and precision. The thoroughness of these results supports the Phoenix drone's potential applications in both research and pedagogical contexts.
Implications and Potential
The open-source nature of the Phoenix drone invites widespread use and innovation from researchers and educators. The platform’s availability of simulation tools and software-in-the-loop (SITL) support furthers accessibility, facilitating enhanced learning experiences and experimental setups without the initial need for physical components.
For the aerial robotics research community, the Phoenix drone provides a versatile platform for exploring advanced control algorithms, cooperative flight dynamics, and vision-based navigation. The educational field can leverage this platform for hands-on student engagement, offering insights into aerodynamics, state estimation, and embedded system design.
Conclusion
The introduction of the Phoenix drone represents a noteworthy contribution to the domain of MAV research and education. The comprehensively detailed open-source resources allow for extensive experimentation and model verification, laying the groundwork for future developments in UAV technology. Students and researchers are well-positioned to benefit from the modular and accessible nature of the Phoenix platform. Future work, potentially exploring enhancements in coordinated multi-drone dynamics or autonomous flight behaviors, could build directly upon this well-developed and documented system.