CrazyMARL: ケーブル吊り下げペイロードの協調輸送のための分散型直接モーター制御ポリシー
CrazyMARL: Decentralized Direct Motor Control Policies for Cooperative Aerial Transport of Cable-Suspended Payloads
複数UAVがケーブルで吊り下げたペイロードを協調輸送するため、分散型強化学習で直接モーター制御ポリシーを学習し、ケーブルの緩み・張り遷移や外乱に頑健で、シミュレーションから実機へのゼロショット転移も実現した。
詳しい要約
1. どんなもの？
2. 先行研究と比べてどこがすごい？
3. 技術・手法の肝は？
4. どうやって有効だと検証した？
5. 議論はある？
6. 次に読むべき論文は？
※ AIが要旨から生成した要約です。正確性は原文をご確認ください。
著者: Viktor Lorentz, Khaled Wahba, Sayantan Auddy, Marc Toussaint, Wolfgang Hönig
分類: cs.RO, cs.MA
原文アブストラクト
Collaborative transportation of cable-suspended payloads by teams of UAVs has the potential to enhance payload capacity, adapt to different payload shapes, and provide built-in compliance, making it attractive for applications ranging from disaster relief to precision logistics. However, multi-UAV coordination under disturbances, nonlinear payload dynamics, and slack-taut cable modes remains a challenging control problem. To our knowledge, no prior work has addressed these cable mode transitions in the multi-UAV context, instead relying on simplifying rigid-link assumptions. We propose CrazyMARL, a decentralized RL framework for multi-UAV cable-suspended payload transport. Simulation results demonstrate that the learned policies can outperform classical decentralized controllers in terms of disturbance rejection and tracking precision, achieving an 80% recovery rate from harsh conditions compared to 44% for the baseline method. We also achieve successful zero-shot sim-to-real transfer and demonstrate that our policies are highly robust under harsh conditions, including wind, random external disturbances, and transitions between slack and taut cable dynamics. This work paves the way for autonomous, resilient UAV teams capable of executing complex payload missions in unstructured environments. Code and videos can be found on the website: https://imrclab.github.io/CrazyMARL.
