Concurrent Learning for Cooperative UAV Transportation of Unknown Payloads

In this work, the transportation problem is addressed by directly attaching the payload to a team of unmanned aerial vehicles (UAVs). The proposed flight controller for cooperative transportation offers a solution by eliminating the need for prior knowledge of payload details, such as the center of gravity (CoG), mass, and moment of inertia (MoI). Typically, the formation for transporting the payload is evenly distributed along the payload boundary. However, this formation can lead to inefficiencies, especially when the CoG of the system is not aligned with the geometric center of the system. In such circumstances, it can result in steady-state error and shorter endurance. The developed controller incorporates a concurrent learning estimator to estimate the mass and CoG simultaneously during flight. This estimation is leveraged to balance power consumption among all UAV agents, resulting in a significant extension of flight time. The system's stability is mathematically proven through the Lyapunov theorem, ensuring a reliable combination of the estimator and adaptive controller. To validate the performance and effectiveness of the proposed approach, simulations and real-world experiments have been conducted, demonstrating the controller's capability to enhance cooperative transportation operations. The results highlight its potential to improve the field of UAV-based payload transportation and provide more efficient and cost-effective transport solutions.