Moment analysis of stochastic hybrid systems using semidefinite programming

Khem Raj Ghusinga, Andrew Lamperski, Abhyudai Singh

Research output: Contribution to journalArticle

Abstract

This paper proposes a method based on semidefinite programming for estimating moments of stochastic hybrid systems (SHSs). The class of SHSs considered herein consists of a finite number of discrete states and a continuous state whose dynamics as well as the reset maps and transition intensities are polynomial in the continuous state. For these SHSs, the dynamics of moments evolve according to a system of linear ordinary differential equations. However, it is generally not possible to exactly solve the system since time evolution of a specific moment may depend upon moments of order higher than it. Our methodology recasts an SHS with multiple discrete modes to a single-mode SHS with algebraic constraints. We then find lower and upper bounds on a moment of interest via a semidefinite program that includes linear constraints obtained from moment dynamics and those arising from the recasting process, along with semidefinite constraints coming from the non-negativity of moment matrices. We illustrate the methodology via an example of SHS.

Original languageEnglish (US)
Article number108634
JournalAutomatica
Volume112
DOIs
StatePublished - Feb 2020

Fingerprint

Hybrid systems
Ordinary differential equations
Polynomials

Keywords

  • Convex programming
  • Jump process
  • Optimization
  • Polynomial models
  • Stochastic systems

Cite this

Moment analysis of stochastic hybrid systems using semidefinite programming. / Ghusinga, Khem Raj; Lamperski, Andrew; Singh, Abhyudai.

In: Automatica, Vol. 112, 108634, 02.2020.

Research output: Contribution to journalArticle

@article{f2ded0ea2df84de88d30992906d6dfab,
title = "Moment analysis of stochastic hybrid systems using semidefinite programming",
abstract = "This paper proposes a method based on semidefinite programming for estimating moments of stochastic hybrid systems (SHSs). The class of SHSs considered herein consists of a finite number of discrete states and a continuous state whose dynamics as well as the reset maps and transition intensities are polynomial in the continuous state. For these SHSs, the dynamics of moments evolve according to a system of linear ordinary differential equations. However, it is generally not possible to exactly solve the system since time evolution of a specific moment may depend upon moments of order higher than it. Our methodology recasts an SHS with multiple discrete modes to a single-mode SHS with algebraic constraints. We then find lower and upper bounds on a moment of interest via a semidefinite program that includes linear constraints obtained from moment dynamics and those arising from the recasting process, along with semidefinite constraints coming from the non-negativity of moment matrices. We illustrate the methodology via an example of SHS.",
keywords = "Convex programming, Jump process, Optimization, Polynomial models, Stochastic systems",
author = "Ghusinga, {Khem Raj} and Andrew Lamperski and Abhyudai Singh",
year = "2020",
month = "2",
doi = "10.1016/j.automatica.2019.108634",
language = "English (US)",
volume = "112",
journal = "Automatica",
issn = "0005-1098",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Moment analysis of stochastic hybrid systems using semidefinite programming

AU - Ghusinga, Khem Raj

AU - Lamperski, Andrew

AU - Singh, Abhyudai

PY - 2020/2

Y1 - 2020/2

N2 - This paper proposes a method based on semidefinite programming for estimating moments of stochastic hybrid systems (SHSs). The class of SHSs considered herein consists of a finite number of discrete states and a continuous state whose dynamics as well as the reset maps and transition intensities are polynomial in the continuous state. For these SHSs, the dynamics of moments evolve according to a system of linear ordinary differential equations. However, it is generally not possible to exactly solve the system since time evolution of a specific moment may depend upon moments of order higher than it. Our methodology recasts an SHS with multiple discrete modes to a single-mode SHS with algebraic constraints. We then find lower and upper bounds on a moment of interest via a semidefinite program that includes linear constraints obtained from moment dynamics and those arising from the recasting process, along with semidefinite constraints coming from the non-negativity of moment matrices. We illustrate the methodology via an example of SHS.

AB - This paper proposes a method based on semidefinite programming for estimating moments of stochastic hybrid systems (SHSs). The class of SHSs considered herein consists of a finite number of discrete states and a continuous state whose dynamics as well as the reset maps and transition intensities are polynomial in the continuous state. For these SHSs, the dynamics of moments evolve according to a system of linear ordinary differential equations. However, it is generally not possible to exactly solve the system since time evolution of a specific moment may depend upon moments of order higher than it. Our methodology recasts an SHS with multiple discrete modes to a single-mode SHS with algebraic constraints. We then find lower and upper bounds on a moment of interest via a semidefinite program that includes linear constraints obtained from moment dynamics and those arising from the recasting process, along with semidefinite constraints coming from the non-negativity of moment matrices. We illustrate the methodology via an example of SHS.

KW - Convex programming

KW - Jump process

KW - Optimization

KW - Polynomial models

KW - Stochastic systems

UR - http://www.scopus.com/inward/record.url?scp=85074158719&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85074158719&partnerID=8YFLogxK

U2 - 10.1016/j.automatica.2019.108634

DO - 10.1016/j.automatica.2019.108634

M3 - Article

AN - SCOPUS:85074158719

VL - 112

JO - Automatica

JF - Automatica

SN - 0005-1098

M1 - 108634

ER -