Reinforcement Learning with Formal Performance Metrics for Quadcopter Attitude Control under Non-nominal Contexts
This work addresses practical controller design for quadcopters in real-world scenarios, though it is incremental in applying existing RL methods to specific non-nominal contexts.
The paper tackled quadcopter attitude control under non-nominal conditions like motor failures and wind gusts by using reinforcement learning with signal temporal logic for performance evaluation, resulting in robust controllers that handle these disturbances effectively.
We explore the reinforcement learning approach to designing controllers by extensively discussing the case of a quadcopter attitude controller. We provide all details allowing to reproduce our approach, starting with a model of the dynamics of a crazyflie 2.0 under various nominal and non-nominal conditions, including partial motor failures and wind gusts. We develop a robust form of a signal temporal logic to quantitatively evaluate the vehicle's behavior and measure the performance of controllers. The paper thoroughly describes the choices in training algorithms, neural net architecture, hyperparameters, observation space in view of the different performance metrics we have introduced. We discuss the robustness of the obtained controllers, both to partial loss of power for one rotor and to wind gusts and finish by drawing conclusions on practical controller design by reinforcement learning.