Pith. sign in

REVIEW

Data-Driven Yet Formal Policy Synthesis for Stochastic Nonlinear Dynamical Systems

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2501.01191 v3 pith:Q3TDH2OD submitted 2025-01-02 eess.SY cs.SY

Data-Driven Yet Formal Policy Synthesis for Stochastic Nonlinear Dynamical Systems

classification eess.SY cs.SY
keywords nonlinearabstractionabstractionscontroldynamicaldynamicsstochasticsynthesis
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

The automated synthesis of control policies for stochastic dynamical systems presents significant challenges. A standard approach is to construct a finite-state abstraction of the continuous system, typically represented as a Markov decision process (MDP). However, generating abstractions is challenging when (1) the system's dynamics are nonlinear, and/or (2) we do not have complete knowledge of the dynamics. In this work, we introduce a novel data-driven abstraction technique for nonlinear Lipschitz continuous dynamical systems with additive stochastic noise that addresses both of these issues. As a key step, we use samples of the dynamics to learn the enabled actions and transition probabilities of the abstraction. We represent abstractions as MDPs with intervals of transition probabilities, known as interval MDPs (IMDPs). These abstractions enable the synthesis of policies for the concrete nonlinear system, with probably approximately correct (PAC) guarantees on the probability of satisfying a specified control objective. Our numerical experiments illustrate the effectiveness and robustness of our approach in achieving reliable control under uncertainty.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.