Certified Stochastic Control via Covariance Steering with Pick-to-Learn

Chun-Wei Kong; Zachary Donovan; Morteza Lahijanian; Jay McMahon

Certified Stochastic Control via Covariance Steering with Pick-to-Learn

Systems and Control 2026-07-23 v1

Authors: Chun-Wei Kong , Zachary Donovan , Morteza Lahijanian , Jay McMahon

Abstract

We present CS-P2L, a framework coupling covariance steering (CS) with the Pick-to-Learn (P2L) meta-algorithm for certified controller synthesis over high-fidelity stochastic simulators. The method iteratively evaluates policies on simulator rollouts, tightens surrogate constraints using the worst-case violations, and provides compression-based probabilistic guarantees on the true violation probability given a confidence level. On a spacecraft powered-descent problem with uncertain gravity, CS-P2L certifies a violation bound of 4.9\% with 600 rollouts, whereas standalone covariance steering underestimates the violation rate by roughly a factor of two.

Keywords

covariance steering certified control stochastic systems Pick-to-Learn probabilistic guarantees

Cite

@article{arxiv.2607.21086,
  title  = {Certified Stochastic Control via Covariance Steering with Pick-to-Learn},
  author = {Chun-Wei Kong and Zachary Donovan and Morteza Lahijanian and Jay McMahon},
  journal= {arXiv preprint arXiv:2607.21086},
  year   = {2026}
}

Comments

To appear in the 65th IEEE Conference on Decision and Control