Active Sampling for Closed-Loop Statistical Verification of Uncertain Nonlinear Systems

John F. Quindlen, Ufuk Topcu, Girish Chowdhary, Jonathan P. How

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Increasingly demanding performance requirements for dynamical systems motivate the adoption of nonlinear and adaptive control techniques. One challenge is the nonlinearity of the resulting closed-loop system complicates verification that the system does satisfy the requirements at all possible operating conditions. This paper presents a data-driven procedure for efficient simulation-based, statistical verification without the reliance upon exhaustive simulations. In contrast to previous work, this approach introduces a method for online estimation of prediction accuracy without the use of external validation sets. This work also develops a novel active sampling algorithm that iteratively selects additional training points in order to maximize the accuracy of the predictions while still limited to a sample budget. Three case studies demonstrate the utility of the new approach and the results show up to a 50% improvement over state-of-the-art techniques.

Original languageEnglish (US)
Title of host publication2018 Annual American Control Conference, ACC 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages6259-6265
Number of pages7
ISBN (Print)9781538654286
DOIs
StatePublished - Aug 9 2018
Event2018 Annual American Control Conference, ACC 2018 - Milwauke, United States
Duration: Jun 27 2018Jun 29 2018

Publication series

NameProceedings of the American Control Conference
Volume2018-June
ISSN (Print)0743-1619

Other

Other2018 Annual American Control Conference, ACC 2018
CountryUnited States
CityMilwauke
Period6/27/186/29/18

Fingerprint

Nonlinear systems
Sampling
Closed loop systems
Dynamical systems

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Cite this

Quindlen, J. F., Topcu, U., Chowdhary, G., & How, J. P. (2018). Active Sampling for Closed-Loop Statistical Verification of Uncertain Nonlinear Systems. In 2018 Annual American Control Conference, ACC 2018 (pp. 6259-6265). [8431662] (Proceedings of the American Control Conference; Vol. 2018-June). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.23919/ACC.2018.8431662

Active Sampling for Closed-Loop Statistical Verification of Uncertain Nonlinear Systems. / Quindlen, John F.; Topcu, Ufuk; Chowdhary, Girish; How, Jonathan P.

2018 Annual American Control Conference, ACC 2018. Institute of Electrical and Electronics Engineers Inc., 2018. p. 6259-6265 8431662 (Proceedings of the American Control Conference; Vol. 2018-June).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Quindlen, JF, Topcu, U, Chowdhary, G & How, JP 2018, Active Sampling for Closed-Loop Statistical Verification of Uncertain Nonlinear Systems. in 2018 Annual American Control Conference, ACC 2018., 8431662, Proceedings of the American Control Conference, vol. 2018-June, Institute of Electrical and Electronics Engineers Inc., pp. 6259-6265, 2018 Annual American Control Conference, ACC 2018, Milwauke, United States, 6/27/18. https://doi.org/10.23919/ACC.2018.8431662
Quindlen JF, Topcu U, Chowdhary G, How JP. Active Sampling for Closed-Loop Statistical Verification of Uncertain Nonlinear Systems. In 2018 Annual American Control Conference, ACC 2018. Institute of Electrical and Electronics Engineers Inc. 2018. p. 6259-6265. 8431662. (Proceedings of the American Control Conference). https://doi.org/10.23919/ACC.2018.8431662
Quindlen, John F. ; Topcu, Ufuk ; Chowdhary, Girish ; How, Jonathan P. / Active Sampling for Closed-Loop Statistical Verification of Uncertain Nonlinear Systems. 2018 Annual American Control Conference, ACC 2018. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 6259-6265 (Proceedings of the American Control Conference).
@inproceedings{2793c9776cb149f4adbb0bc01266457b,
title = "Active Sampling for Closed-Loop Statistical Verification of Uncertain Nonlinear Systems",
abstract = "Increasingly demanding performance requirements for dynamical systems motivate the adoption of nonlinear and adaptive control techniques. One challenge is the nonlinearity of the resulting closed-loop system complicates verification that the system does satisfy the requirements at all possible operating conditions. This paper presents a data-driven procedure for efficient simulation-based, statistical verification without the reliance upon exhaustive simulations. In contrast to previous work, this approach introduces a method for online estimation of prediction accuracy without the use of external validation sets. This work also develops a novel active sampling algorithm that iteratively selects additional training points in order to maximize the accuracy of the predictions while still limited to a sample budget. Three case studies demonstrate the utility of the new approach and the results show up to a 50{\%} improvement over state-of-the-art techniques.",
author = "Quindlen, {John F.} and Ufuk Topcu and Girish Chowdhary and How, {Jonathan P.}",
year = "2018",
month = "8",
day = "9",
doi = "10.23919/ACC.2018.8431662",
language = "English (US)",
isbn = "9781538654286",
series = "Proceedings of the American Control Conference",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "6259--6265",
booktitle = "2018 Annual American Control Conference, ACC 2018",
address = "United States",

}

TY - GEN

T1 - Active Sampling for Closed-Loop Statistical Verification of Uncertain Nonlinear Systems

AU - Quindlen, John F.

AU - Topcu, Ufuk

AU - Chowdhary, Girish

AU - How, Jonathan P.

PY - 2018/8/9

Y1 - 2018/8/9

N2 - Increasingly demanding performance requirements for dynamical systems motivate the adoption of nonlinear and adaptive control techniques. One challenge is the nonlinearity of the resulting closed-loop system complicates verification that the system does satisfy the requirements at all possible operating conditions. This paper presents a data-driven procedure for efficient simulation-based, statistical verification without the reliance upon exhaustive simulations. In contrast to previous work, this approach introduces a method for online estimation of prediction accuracy without the use of external validation sets. This work also develops a novel active sampling algorithm that iteratively selects additional training points in order to maximize the accuracy of the predictions while still limited to a sample budget. Three case studies demonstrate the utility of the new approach and the results show up to a 50% improvement over state-of-the-art techniques.

AB - Increasingly demanding performance requirements for dynamical systems motivate the adoption of nonlinear and adaptive control techniques. One challenge is the nonlinearity of the resulting closed-loop system complicates verification that the system does satisfy the requirements at all possible operating conditions. This paper presents a data-driven procedure for efficient simulation-based, statistical verification without the reliance upon exhaustive simulations. In contrast to previous work, this approach introduces a method for online estimation of prediction accuracy without the use of external validation sets. This work also develops a novel active sampling algorithm that iteratively selects additional training points in order to maximize the accuracy of the predictions while still limited to a sample budget. Three case studies demonstrate the utility of the new approach and the results show up to a 50% improvement over state-of-the-art techniques.

UR - http://www.scopus.com/inward/record.url?scp=85052605200&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85052605200&partnerID=8YFLogxK

U2 - 10.23919/ACC.2018.8431662

DO - 10.23919/ACC.2018.8431662

M3 - Conference contribution

AN - SCOPUS:85052605200

SN - 9781538654286

T3 - Proceedings of the American Control Conference

SP - 6259

EP - 6265

BT - 2018 Annual American Control Conference, ACC 2018

PB - Institute of Electrical and Electronics Engineers Inc.

ER -