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.
Turn this paper into a full lesson
ArcXiv compiles a staged curriculum from this paper: 8-12 lessons across beginner → advanced, synthesised section guides, visuals, flashcards, a quiz, exercises, and on-demand deep dives per section. Grounded in the abstract, never invented.