ePrints

Formal Verification of Unknown Discrete- and Continuous-Time Systems: A Data-Driven Approach

Lookup NU author(s): Dr Amy Nejati, Dr Abolfazl Lavaei, Dr Sadegh SoudjaniORCiD

Downloads

Full text for this publication is not currently held within this repository. Alternative links are provided below where available.

Abstract

This paper is concerned with a formal verification scheme for both discrete- and continuous-time deterministic systems with unknown mathematical models. The main target is to verify the safety of unknown systems based on the construction of barrier certificates via a set of data collected from trajectories of systems while providing an a-priori guaranteed confidence on the safety. In our proposed framework, we first cast the original safety problem as a robust convex program (RCP). Solving the proposed RCP is not tractable in general since the unknown model appears in one of the constraints. Instead, we collect finite numbers of data from trajectories of the system and provide a scenario convex program (SCP) corresponding to the original RCP. We then establish a probabilistic closeness between the optimal value of SCP and that of RCP, and as a result, we formally quantify the safety guarantee of unknown systems based on the number of data points and a required level of confidence. We propose our framework in both discrete-time and continuous-time settings. We illustrate the effectiveness of our proposed results by first applying them to an unknown continuous-time room temperature system. We verify that the temperature of the room maintains in a comfort zone with some desirable confidence by collecting data from trajectories of the system. To show the applicability of our techniques to higher dimensional systems with nonlinear dynamics, we then apply our results to a continuous-time nonlinear jet engine compressor and a discrete-time DC motor.

Publication metadata

Author(s): Nejati A, Lavaei A, Jagtap P, Soudjani S, Zamani M

Publication type: Article

Publication status: Published

Journal: IEEE Transactions on Automatic Control

Year: 2023

Volume: 68

Issue: 5

Pages: 3011-3024

Print publication date: 01/05/2023

Online publication date: 10/03/2023

Acceptance date: 11/02/2023

ISSN (print): 0018-9286

ISSN (electronic): 1558-2523

Publisher: IEEE

URL: https://doi.org/10.1109/TAC.2023.3255141

DOI: 10.1109/TAC.2023.3255141

Altmetrics

Share