DR-PETS: Learning-Based Control With Planning in Adversarial Environments
This work addresses robustness in control systems for real-world decision-making, but it is incremental as it builds upon the existing PETS algorithm.
The paper tackled the problem of ensuring robustness against adversarial perturbations in learning-based control by proposing DR-PETS, a distributionally robust extension of PETS, which certifies robustness and achieves consistent performance in worst-case scenarios where PETS deteriorates, as demonstrated in pendulum stabilization and cart-pole balancing experiments.
Ensuring robustness against epistemic, possibly adversarial, perturbations is essential for reliable real-world decision-making. While the Probabilistic Ensembles with Trajectory Sampling (PETS) algorithm inherently handles uncertainty via ensemble-based probabilistic models, it lacks guarantees against structured adversarial or worst-case uncertainty distributions. To address this, we propose DR-PETS, a distributionally robust extension of PETS that certifies robustness against adversarial perturbations. We formalize uncertainty via a p-Wasserstein ambiguity set, enabling worst-case-aware planning through a min-max optimization framework. While PETS passively accounts for stochasticity, DR-PETS actively optimizes robustness via a tractable convex approximation integrated into PETS planning loop. Experiments on pendulum stabilization and cart-pole balancing show that DR-PETS certifies robustness against adversarial parameter perturbations, achieving consistent performance in worst-case scenarios where PETS deteriorates.