Integrated Passive-Active Model Identification with Tunable Model Discrimination for Affine Discrete-Time Systems

Changrui Liu; Qiang Shen; Ruochen Niu; Sze Zheng Yong

doi:10.1109/LCSYS.2021.3135029

Integrated Passive-Active Model Identification with Tunable Model Discrimination for Affine Discrete-Time Systems

Changrui Liu, Qiang Shen, Ruochen Niu, Sze Zheng Yong

Engineering, Ira A. Fulton Schools of (IAFSE)

Research output: Contribution to journal › Article › peer-review

Abstract

This letter proposes a passive-active model identification algorithm for affine discrete-time systems that integrates active model discrimination (AMD) and model invalidation (MI). A look-up tree consisting of control inputs is constructed offline for this integrated model identification (IMI) technique to discriminate among models in a time-varying model set, which is only known at run time when repeatedly applying MI online. Furthermore, a novel tunable AMD (TAMD), with its mixed-integer linear programming (MILP) formulation, is proposed and combined with the IMI algorithm, which can improve model discrimination performance. The effectiveness of the proposed IMI algorithm is demonstrated through simulations for identifying intention models of human-driven vehicles in a lane changing scenario.

Original language	English (US)
Pages (from-to)	1885-1890
Number of pages	6
Journal	IEEE Control Systems Letters
Volume	6
DOIs	https://doi.org/10.1109/LCSYS.2021.3135029
State	Published - 2022

Keywords

Model validation
estimation
fault diagnosis

ASJC Scopus subject areas

Control and Systems Engineering
Control and Optimization

Access to Document

10.1109/LCSYS.2021.3135029

Cite this

@article{0079f578a71c46d993c7ae88b9c5cff4,

title = "Integrated Passive-Active Model Identification with Tunable Model Discrimination for Affine Discrete-Time Systems",

abstract = "This letter proposes a passive-active model identification algorithm for affine discrete-time systems that integrates active model discrimination (AMD) and model invalidation (MI). A look-up tree consisting of control inputs is constructed offline for this integrated model identification (IMI) technique to discriminate among models in a time-varying model set, which is only known at run time when repeatedly applying MI online. Furthermore, a novel tunable AMD (TAMD), with its mixed-integer linear programming (MILP) formulation, is proposed and combined with the IMI algorithm, which can improve model discrimination performance. The effectiveness of the proposed IMI algorithm is demonstrated through simulations for identifying intention models of human-driven vehicles in a lane changing scenario.",

keywords = "Model validation, estimation, fault diagnosis",

author = "Changrui Liu and Qiang Shen and Ruochen Niu and Yong, {Sze Zheng}",

note = "Funding Information: This work was supported in part by the Natural Science Foundation of China under Grant 62103275, and in part by the National Science Foundation under Grant CNS-1943545. Publisher Copyright: {\textcopyright} 2017 IEEE.",

year = "2022",

doi = "10.1109/LCSYS.2021.3135029",

language = "English (US)",

volume = "6",

pages = "1885--1890",

journal = "IEEE Control Systems Letters",

issn = "2475-1456",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - Integrated Passive-Active Model Identification with Tunable Model Discrimination for Affine Discrete-Time Systems

AU - Liu, Changrui

AU - Shen, Qiang

AU - Niu, Ruochen

AU - Yong, Sze Zheng

N1 - Funding Information: This work was supported in part by the Natural Science Foundation of China under Grant 62103275, and in part by the National Science Foundation under Grant CNS-1943545. Publisher Copyright: © 2017 IEEE.

PY - 2022

Y1 - 2022

N2 - This letter proposes a passive-active model identification algorithm for affine discrete-time systems that integrates active model discrimination (AMD) and model invalidation (MI). A look-up tree consisting of control inputs is constructed offline for this integrated model identification (IMI) technique to discriminate among models in a time-varying model set, which is only known at run time when repeatedly applying MI online. Furthermore, a novel tunable AMD (TAMD), with its mixed-integer linear programming (MILP) formulation, is proposed and combined with the IMI algorithm, which can improve model discrimination performance. The effectiveness of the proposed IMI algorithm is demonstrated through simulations for identifying intention models of human-driven vehicles in a lane changing scenario.

AB - This letter proposes a passive-active model identification algorithm for affine discrete-time systems that integrates active model discrimination (AMD) and model invalidation (MI). A look-up tree consisting of control inputs is constructed offline for this integrated model identification (IMI) technique to discriminate among models in a time-varying model set, which is only known at run time when repeatedly applying MI online. Furthermore, a novel tunable AMD (TAMD), with its mixed-integer linear programming (MILP) formulation, is proposed and combined with the IMI algorithm, which can improve model discrimination performance. The effectiveness of the proposed IMI algorithm is demonstrated through simulations for identifying intention models of human-driven vehicles in a lane changing scenario.

KW - Model validation

KW - estimation

KW - fault diagnosis

UR - http://www.scopus.com/inward/record.url?scp=85122156414&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85122156414&partnerID=8YFLogxK

U2 - 10.1109/LCSYS.2021.3135029

DO - 10.1109/LCSYS.2021.3135029

M3 - Article

AN - SCOPUS:85122156414

SN - 2475-1456

VL - 6

SP - 1885

EP - 1890

JO - IEEE Control Systems Letters

JF - IEEE Control Systems Letters

ER -

Integrated Passive-Active Model Identification with Tunable Model Discrimination for Affine Discrete-Time Systems

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this