Hysteresis-Based RL: Robustifying Reinforcement Learning-based Control Policies via Hybrid Control
This work addresses robustness issues in RL control policies for complex systems, which is an incremental improvement over existing methods.
The authors tackled the lack of robustness guarantees in reinforcement learning-based control policies by proposing Hysteresis-Based RL (HyRL), a hybrid algorithm that augments existing RL methods with hysteresis switching and two-stage learning, demonstrating its effectiveness in examples where PPO and DQN fail.
Reinforcement learning (RL) is a promising approach for deriving control policies for complex systems. As we show in two control problems, the derived policies from using the Proximal Policy Optimization (PPO) and Deep Q-Network (DQN) algorithms may lack robustness guarantees. Motivated by these issues, we propose a new hybrid algorithm, which we call Hysteresis-Based RL (HyRL), augmenting an existing RL algorithm with hysteresis switching and two stages of learning. We illustrate its properties in two examples for which PPO and DQN fail.