AcroRL: Learning Aggressive Quadrotor Inversion using Bidirectional Thrust
AcroRL: Learning Aggressive Quadrotor Inversion using Bidirectional Thrust
This paper introduces AcroRL, a learning-based framework that enables quadrotors with bidirectional thrust to perform aggressive maneuvers such as inverted flight and perching. By training separate reinforcement learning policies for transitions between nominal and inverted states, the method achieves lower position RMSE (32% reduction) and settling time (57% reduction) compared to optimization-based baselines in simulation. Hardware experiments confirm successful inversion across multiple yaw configurations with position RMSE below 0.35m, while maintaining compatibility with traditional trajectory generation. An open-source implementation is provided.
The study presents a reinforcement learning framework for quadrotors with bidirectional thrust, demonstrating reduced position RMSE and settling time in simulation alongside hardware validation of inversion capabilities.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- SimulationOpen source S4
- paper: not_reported
- weights: not_reported
- dataset: not_reported
- code: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
ed via reinforcement learning for nominal-to-inverted and inverted-to-nominal transitions. In JAX-based simulation, the proposed method achieves the lowest position deviation and settling time across all evaluated baselines, reducing position root mean square error (RMSE) by 32% and settling time by 57% relative to the strongest optimization-based baseline. Hardware experiments demonstrate successful inversion across
Open source S4
multiple yaw configurations with position RMSE below 0.35m, and compatibility with downstream trajectory generation and control through circular flight in both regimes. Additionally, we provide an open-source implementation of the proposed framework.
Open source S5
Source：arXiv Robotics — research abstracts · arxiv.org