arXiv · 2603.21123
VisFly-Lab: Unified Differentiable Framework for First-Order Reinforcement Learning of Quadrotor Control
Abstract
First-order reinforcement learning with differentiable simulation is promising for quadrotor control, but practical progress remains fragmented across task-specific settings. To support more systematic development and evaluation, we present a unified differentiable framework for multi-task quadrotor control. The framework is wrapped, extensible, and equipped with deployment-oriented dynamics, providing a common interface across four representative tasks: hovering, tracking, landing, and racing. We also present the suite of first-order learning algorithms, where we identify two practical bottlenecks of standard first-order training: limited state coverage caused by horizon initialization and gradient bias caused by partially non-differentiable rewards. To address these issues, we propose Amended Backpropagation Through Time (ABPT), which combines differentiable rollout optimization, a value-based auxiliary objective, and visited-state initialization to improve training robustness. Experimental results show that ABPT yields the clearest gains in tasks with partially non-differentiable rewards, while remaining competitive in fully differentiable settings. We further provide proof-of-concept real-world deployments showing initial transferability of policies learned in the proposed framework beyond simulation.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Fanxing Li, Fangyu Sun, Tianbao Zhang, Shuyu Wu, Dexin Zuo, yufei Yan, Wenxian Yu, Danping Zou. 2026-03-22. VisFly-Lab: Unified Differentiable Framework for First-Order Reinforcement Learning of Quadrotor Control. https://arxiv.org/abs/2603.21123
Cite the original work for its findings. Save a collection to share your selection of sources.