Sim-to-(Multi)-Real: Transfer of Low-Level Robust Control Policies to Multiple Quadrotors
This work addresses the need for robust, generalizable low-level control in quadrotors, which is incremental as it builds on existing sim-to-real transfer methods but demonstrates novel generalization to multiple platforms.
The authors tackled the problem of quadrotor stabilizing controllers requiring model-specific tuning by using reinforcement learning to train low-level policies in simulation that transfer robustly to multiple physical quadrotors, with results showing policies that withstand disturbances and harsh initial conditions like throws.
Quadrotor stabilizing controllers often require careful, model-specific tuning for safe operation. We use reinforcement learning to train policies in simulation that transfer remarkably well to multiple different physical quadrotors. Our policies are low-level, i.e., we map the rotorcrafts' state directly to the motor outputs. The trained control policies are very robust to external disturbances and can withstand harsh initial conditions such as throws. We show how different training methodologies (change of the cost function, modeling of noise, use of domain randomization) might affect flight performance. To the best of our knowledge, this is the first work that demonstrates that a simple neural network can learn a robust stabilizing low-level quadrotor controller (without the use of a stabilizing PD controller) that is shown to generalize to multiple quadrotors.